Australia performs better than the OECD average on a school equity indicator, and yet its most disadvantaged students have not progressed by a single point in ten years. The gap in science scores between the most advantaged and least advantaged quarters fell to 80 points in 2025, below the OECD average of 85, according to PISA results published by the OECD. But this narrowing is explained almost entirely by a decline among advantaged students, while results for disadvantaged students stagnate. A school that becomes more equal at the bottom does not create social mobility.
The essentials
- In Australia, the gap in science scores between advantaged and disadvantaged students fell below the OECD average, but this result comes from a decline among advantaged students, not progress among disadvantaged ones (OECD, PISA 2025).
- Australian students in the least advantaged quarter have stagnated since 2015: their absolute level has not improved.
- Reducing a relative gap and raising absolute performance are two distinct objectives; confusing them leads to policies that improve statistics without improving skills.
- Countries that combine equity and high performance, such as Canada, Estonia, and Japan, show that it is possible to narrow gaps from the top up rather than from the bottom down.
- The question for Australia by 2035: whether its education policies set targets for absolute progress among disadvantaged students, not just for reducing relative gaps.
A flattering indicator that masks stagnation
Australia is a country that invests heavily in its public education system. Since the 2010s, successive governments have made equity a priority: differentiated funding based on school needs, improvement plans targeting disadvantaged areas, strengthened statistical monitoring. The 2025 PISA results appear to validate this direction: with an 80-point gap between advantaged and disadvantaged students in science, Australia falls below the OECD average.
The problem lies in the mechanism. When a gap narrows, two configurations are possible: students at the bottom progress faster than those at the top, or students at the top decline while those at the bottom stagnate. The first produces mobility. The second produces a statistic.
In Australia since 2015, it is the second that is observed. Scores for students in the most advantaged quarter have declined. Those for students in the least advantaged quarter have not moved. The gap has narrowed, but on average, students in the lower socio-economic quarter registered no statistically significant progress between 2015 and 2025. What dashboards present as progress in equity is, upon closer inspection, leveling down.
Several English-speaking countries have experienced comparable trajectories since PISA 2012, and Australia offers the most documented illustration. The OECD country note records both the narrowing of the gap between advantaged and disadvantaged students and the stagnation of absolute levels—two elements rarely read together.
PISA: what the survey measures and how its findings are used
PISA measures the skills of 15-year-olds in reading, mathematics, and science, every three years, across roughly sixty countries. The survey provides two types of information: absolute levels of skills, and gaps between student groups. These two pieces of information answer two different questions.
Absolute level indicates whether students can read, calculate, and reason sufficiently to enter the job market, pursue further studies, and adapt. The gap indicates whether the school system treats students fairly regardless of their social background.
These two questions are legitimate. They can even be answered simultaneously: countries that raise their average level while reducing their gaps exist. But they can also diverge, and that is precisely what the Australian data shows. A school system can narrow its gaps by pulling down the top without touching the bottom. It becomes more equal without becoming more effective.
The confusion between the two is as much political as statistical. When a government announces a reduction in school inequality, media and public opinion tend to read: disadvantaged students are progressing. The OECD publishes both data series, but the gap indicator is simpler to communicate than a breakdown of trajectory by group. The result: public policies driven by an indicator that can improve without children having actually learned anything more.
Australia is not alone in steering by this indicator. But it is a useful case precisely because the 2025 PISA data allow the two effects to be clearly separated. What the Australian economy struggles to produce in terms of productive renewal may perhaps find a root in this difficulty in raising the level of real competencies of its working population.
Social mobility requires skills, not relative equality
Social mobility is the result of a concrete process: a child born into a modest background acquires skills that enable them to access jobs and incomes higher than those of their parents. This process goes through school, but it is not automatic. It assumes the school actually transmits these skills, not simply distributes them more equally.
A student whose science score stagnates at 420 PISA points between 2015 and 2025, while their advantaged classmate declines from 530 to 500 points, finds themselves in a statistically more favorable position relative to that classmate. Yet they remain at 420 points. These 420 points do not allow them to apply for a selective science program, nor read a technical manual with ease, nor understand the probabilities that structure a medical or financial decision. The gap has narrowed; their life trajectory has not changed.
This is what education economists call the difference between relative equity and absolute progress. One is measured between groups. The other is measured against a competency threshold that conditions real opportunities. The two indicators can move in opposite directions, and that is precisely what is observed in Australia.
Countries that have genuinely raised social mobility over the past thirty years have set their education policies on absolute progress for the most vulnerable students, not simply on reducing gaps. Finland in the 1970s-1990s, Poland in the 2000s, Estonia and Vietnam more recently, have reduced internal inequalities while raising their competency floor. Their disadvantaged students progressed in absolute terms, not merely relative to declining advantaged peers.
The decline of top-performing students in Australia
Understanding the decline of advantaged Australian students is as important as documenting the stagnation of disadvantaged ones. The two phenomena can have common or distinct causes.
Several factors are documented in comparative PISA analyses. Actual learning time has declined in several English-speaking countries since 2010. Screen use in classrooms, which became widespread in Australia before most OECD countries, has been subject to nuanced evaluations: studies conducted on PISA data show a negative correlation between intensive use of digital devices in class and reading and science scores, though causality is not entirely established. Australia began reflecting on this issue from 2023, and several states have begun limiting phone use in schools.
The question of teacher recruitment and retention is also documented. Australia faces a shortage of qualified teachers in science subjects, particularly in rural and peri-urban schools that educate a significant share of advantaged students in less densely populated areas. This point connects to a broader finding on teaching team stability: in the OECD, systems where secondary teachers remain in their positions over the long term show more stable results over time.
The decline at the top and stagnation at the bottom also share the fact that they have not triggered sufficient political alarm, precisely because the gap between the two groups was narrowing. A monitoring system that tracks relative gaps without monitoring absolute levels creates blind spots.
Practices of systems that lift all their students
The 2025 PISA data allow Australia to be compared with systems that combine high performance and equity. Canada, Estonia, and Japan regularly appear among these countries. Their trajectories are different, but they share several characteristics.
These systems set their education policy objectives around competency thresholds to reach, not just gap indicators. A disadvantaged Canadian student is tracked on their reading progress against an expected level, not merely relative to wealthier classmates. This difference in policy framing is decisive: it forces system actors to work on absolute level, not to settle for a relative reduction.
These systems invest heavily in initial and ongoing teacher training, targeting pedagogical practices whose effectiveness is documented. Estonia, for example, conducted a curriculum reform from the 2000s onward based on rigorous selection of teaching methods. The result is a high average level with small gaps, without sacrificing one for the other.
Japan combines relatively homogeneous classes in terms of methods with enhanced support for struggling students during and after school hours. The model is difficult to transpose directly into a country as diverse as Australia, but the principle—acting on the level of all students rather than on the gap between them—is transferable.
These examples are not universal recipes. They show that combining high performance and equity is not a contradiction. It simply requires not confusing the two objectives and not being satisfied with one when the other stagnates.
Two trajectories for Australian education by 2035
Australia is at a crossroads. The 2025 PISA data make visible a trend that has been unfolding for ten years. What the system does with this information over the coming years will determine two very different trajectories.
In the first, Australian education policy continues to be steered by relative gap indicators. The gap between advantaged and disadvantaged students narrows further, for the same reasons as between 2015 and 2025: advantaged students stagnate or decline slightly, disadvantaged ones do not progress. Australia improves its ranking on comparative equity tables. It manufactures an appearance of equality without real mobility and without raising the average level. In this scenario, jobs requiring solid skills in science and reading continue to be accessible mainly to students educated in private schools or in the best-resourced areas, independent of gap statistics.
In the second trajectory, the 2025 PISA data serve as a political alarm signal. The federal government and states explicitly set targets for absolute progress for students in the least advantaged quarter, independent of how advantaged students evolve. Funding is directed toward pedagogical practices whose effectiveness is documented in comparable contexts. Teacher training in mathematics and science is strengthened in schools serving the most vulnerable populations. Monitoring indicators distinguish between the two types of gap reduction, and policy makers report on both.
The bifurcation between these two trajectories turns first on a political framing, not on considerable additional resources. Aiming to reduce a relative gap and aiming for absolute progress are two compatible objectives, but the second is more demanding, less flattering in the short term, and more useful to the students involved.
The signals to watch to know which trajectory Australia is on in two or three years are relatively simple: the change in average score for students in the least advantaged quarter in science, independent of the gap with advantaged peers, and the change in average score for the most advantaged quarter, to know whether any tightening comes from decline at the top or progress at the bottom. PISA 2028 will provide this data. Australian policies put in place by then will decide which to observe.
The data invite Australia, like all education systems that measure equity primarily through gaps, to transmit knowledge that grounds social mobility, rather than produce statistics that simulate it.



