Science1 distinct publisher3 min readUpdated
Twenty-four years of HILDA data show the hourly wage advantage of a degree peaked around 2010 and fell to 38% by 2024. The VET premium is down to 2%.
The Scientist · Science desk

Compiled by The ScientistSomething wrong?How this is made
Adelaide University researchers have measured the wage return to post-school education in Australia year by year rather than in aggregate, and found the hourly premium for a degree peaked at about 53% around 2010 before falling to 38% by 2024 [1][4][5]. That matters because the headline figure usually quoted for Australia, roughly 46%, is the 24-year average across the study window and sits eight percentage points above where the premium actually finished [3][7].
The study, published in International Review of Economics & Finance, used more than 250,000 observations of people aged 25 to 64 from the Household, Income and Labour Dynamics in Australia survey between 2001 and 2024 [1][2]. That is roughly 10,400 observations a year, enough to estimate an annual series rather than a period mean [6]. The design choice is the finding. Comparing only 2001 with 2024 shows relatively little net change in hourly premiums; the annual series shows a rise through the early 2000s followed by a prolonged decline [6]. Anyone who has been quoting stable endpoint comparisons has been describing a round trip as a flat line, and the 2024 hourly figure implies the 2001 starting point was in the same neighbourhood as today's [8].
The vocational picture is worse in relative terms. The VET hourly premium peaked at about 14% around 2010 and was 2% by 2024, a fall of 12 percentage points that erases roughly 86% of the advantage [4][5][2]. The degree premium's 15-point fall is a 28% relative decline [1]. In other words, a VET qualification's hourly wage advantage over Year 12 or below has all but disappeared, while a degree's has been trimmed [4][5].
The sharper signal is in hours. On weekly earnings, which combine wages and hours worked, the degree premium peaked at 71% around 2008 and fell to 40% by 2024, a 44% relative decline [7][3]. The VET weekly premium peaked at 32% in 2005 and reached 3% in 2024 [8][4]. At their respective peaks, the weekly premium ran 18 points above the hourly premium for graduates; by 2024 the gap was two points [5]. Lead researcher Dr Tinh Doan attributes that accelerated decline in weekly relative earnings to reduced work hours and underemployment [9]. Graduates used to earn more per hour and work more hours; the second advantage has largely gone [5][9].
Doan says the erosion coincides with rapid growth in post-school graduate numbers, structural shifts toward lower-productivity sectors, and skill mismatch, citing that about 30% of high-skilled workers are now in middle- or low-skilled roles [10]. On his account, returns reflect how qualifications are valued when the supply of skilled workers grows faster than the economy creates high-skilled jobs [11]. The data as described does not separate a devalued credential from a poor job match, and the hours channel points at demand for the work rather than the certificate itself [9][10].
The policy timing is awkward. The Australian Tertiary Education Commission, established in July 2025, supports a national target of 80% of the working-age population holding tertiary qualifications by 2050, 25 years out [12][9]. Doan warns that expanding participation without parallel growth in demand for high-skilled, non-routine work could push relative wages down further and raise graduate underemployment risk [13].
Watch for the follow-up work Doan says is needed on returns by field of study, occupation and cohort, including migrant versus local workers [14]. Until that lands, the aggregate premium tells students little about which course pays [15].
Follow any of these and your For You feed starts watching them — no settings page required.
Ranked by verification strength, evidence, and original report placement.
Adelaide University research published in International Review of Economics & Finance examined how the financial returns to higher education and vocational education and training in Australia changed each year between 2001 and 2024.
Researchers analysed 24 years of data from the Household, Income and Labour Dynamics in Australia (HILDA) survey, covering more than 250,000 observations of people aged 25 to 64.
Across the 2001-2024 period, higher education degree holders earned an average hourly wage premium of roughly 46%, while VET qualifications yielded an average premium of around 10%.
Hourly wage returns rose through the early 2000s and peaked around 2010, when higher education graduates earned about 53% more than people with Year 12 or lower qualifications and VET holders earned about 14% more.
By 2024 the hourly premiums were 38% for higher education graduates and 2% for VET graduates.
Comparing only the endpoints of 2001 and 2024 shows relatively little net change in hourly wage premiums, while the year-by-year analysis found a rise followed by a prolonged decline.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Peer-reviewed panel study, single-outlet reporting
The core numbers rest on a peer-reviewed article in International Review of Economics & Finance built on 24 years of HILDA data and more than 250,000 observations, with a DOI supplied - strong for the descriptive series. It is weakened by having only one publisher (a release-style item), no reported uncertainty ranges or model detail, and by causal and normative layers that the material itself flags as unresolved.
No adoption or uptake evidence supplied
The cluster contains a research finding, not a deployable artefact, and the supplied material records no citations, policy adoption, provider program changes or other measurable uptake. The ATEC target predates and is independent of this study, so it cannot be read as adoption of these findings.
Slightly overstated at the policy edge
The quantitative core is presented conservatively and even discloses that endpoint comparisons would look flat, which is anti-hype. The modest positive gap comes from the causal and forward-looking framing carried without qualification - hours, sectoral shift and mismatch as explanations, and a projection of further wage pressure - plus the implicit critique of the 80% by 2050 target, none of which is tested in the supplied material or accompanied by uncertainty ranges.
Institution promoting its own research, with policy ask
The single item is an Adelaide University release republished by a science-news aggregator: the institution and its lead researcher benefit from visibility, and the quoted conclusions advocate a specific policy posture (align education expansion with high-skilled labour demand) and further research funding. Countervailing factors are the peer-reviewed publication, the use of an independent public dataset, and disclosure of the less favourable endpoint framing.
Solid on numbers, thin on corroboration
Confidence is moderate: the descriptive premium series is precise, internally consistent and drawn from a peer-reviewed analysis of a large national panel, so the direction and magnitude of the decline are likely robust. It is held down by a one-publisher cluster with no independent commentary, absent uncertainty estimates and composition controls, and no adoption evidence at all.
science
What you expect from your own old age shows up a decade later in who you still see1 distinct publisher
science
Eastern US extreme rain is pooling into fewer, wider storms, and station records hide it1 distinct publisher
science
A named ship and a published timetable turn the Arctic into a bookable lane1 distinct publisher
science
Content cues finished last: a 204-sample meta-analysis on workplace first impressions1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
1 article · August 20, 2026