
The graduate employment panic
A small dip in graduate employment and suddenly the media shouts that the robots are coming for the entry level jobs. For students approaching graduation, and for parents quietly recalculating expectations, it feels very real. A small fall in full-time outcomes looks like the beginning of something permanent. Artificial intelligence is advancing quickly and fear of the unknown is everywhere.
But look closely and the pattern is far less dramatic. In Australia and Singapore, graduate employment surged to post-pandemic highs in 2022 and 2023 before easing in 2024. Even after the correction, outcomes remain above pre-COVID levels. In Australia, full-time employment for domestic undergraduates peaked in 2023 and softened in 2024, yet still sits above 2019. Singapore shows a similar shape: a strong rebound in full-time permanent roles followed by a modest pullback. Japan continues to report near-record hiring rates for new graduates. India and China face scale and matching challenges, but not a sudden collapse.
What we are seeing is not the first wave of displacement. It is the labour market normalising after an extraordinary post-COVID rebound.
The context matters. The hiring environment of 2022 and 2023 was unusual. Borders reopened and delayed recruitment resumed. Employers rebuilt pipelines that had been frozen or distorted during the pandemic while graduates entered one of the tightest entry-level labour markets in years. Against that backdrop, 2024 looks weaker, but it is weaker relative to an abnormal peak, not relative to history.
If AI were already displacing graduates at scale, the signal would look different. You would expect persistent declines across multiple cohorts, not a one-year softening. You would expect sharper falls in task-heavy white-collar pathways than in roles requiring judgement, client interaction or physical presence. You would also see divergence between overall labour market conditions and graduate outcomes.
The first effects of automation tend to appear as substitution rather than disappearance. AI changes how work is done before it eliminates the work itself. Entry-level roles evolve and some tasks become automated while others become more analytical or collaborative. The premium shifts toward judgement, adaptability and the ability to work with new tools. The first rung of the career ladder rarely disappears, but it does change shape.
While AI gets the headlines, there is a deeper force at work that receives less attention: demographics. As I highlighted in a recent speech, across much of the developed world, the old-age dependency ratio is rising sharply. Fewer workers are supporting more retirees. In Japan there are now roughly two working-age adults for every person over 65. Across Europe and parts of East Asia, median ages continue to climb. Even Australia and Singapore, while younger, are ageing steadily. When the working-age share of the population declines, economies have only three levers: raise participation, lift productivity or import labour.
Automation sits squarely in the productivity category. It is not simply a cost-saving tool; it is a demographic necessity.
At the same time, demographic momentum is uneven. India continues to add millions of working-age adults each year. Parts of Southeast Asia remain relatively young. China’s graduate cohort continues to grow in absolute terms even as its broader population ages. Over time, digital and analytical work will both be automated and redistributed to where graduate supply exists. Some tasks will be done by software while others will be done by capable young workers in the geographies where they are available. Regardless of location, the premium will increasingly fall on adaptable skills rather than static qualifications.
Seen through this lens, modest fluctuations in graduate employment look less like systemic failure and more like early adjustments in a system responding to ageing populations and shifting talent pools.
For students approaching graduation, and for parents watching from the sidelines, the anxiety is understandable. The headlines are loud. A percentage point here, a hiring cut there, and it is easy to imagine the ladder being pulled up just as the next cohort arrives.
But the evidence does not point to a vanishing first rung. It points to a labour market that surged, then cooled. To employers recalibrating after a hiring rush. To technology reshaping tasks but not eliminating the need for talent. And to a world in which ageing societies will need capable, adaptable graduates more, not less.
The challenge for this generation is not that there will be no jobs. It is that the jobs will look different. Six years ago at Australia’s National Press Club, when launching Deloitte’s report on the future of work, we showed that employers place the highest value on skills that have been acquired recently. Careers will increasingly involve cycles of learning, working and re-skilling throughout life.
Graduate jobs are not disappearing, but the shelf life of skills is getting shorter. That points to entering the workforce sooner rather than accumulating extra degrees, and to placing greater emphasis on lifelong learning. A decade ago I argued that universities would need to adapt to a changing market. The persistent decline in student experience across most tertiary systems suggests that change is not happening quickly enough.