The Skills Gap Is Only Half the Story: Why Graduate Employability Can Erode After University

Universities are under growing pressure to produce graduates who can compete in digitally intensive labour markets, yet the value of what students learn can shift before they ever enter a workplace. Technologies change, job tasks are redesigned and employer expectations move, sometimes faster than formal degree programmes can respond.

A new study, "Digital Skill Alignment over Time: The Dynamic Digital Skill Realisation Framework for Graduate Employability," authored by Ioannis Zervas of the University of Macedonia and published in Education Sciences, introduces the Dynamic Digital Skill Realisation Framework (DDSRF) to explain why graduate employability cannot be reduced to whether universities teach the skills employers currently request.

The framework follows digital capability across the university-to-work transition. It examines how skills are formed, how well they initially match labour-market demand, how their relevance can change, how graduates make them visible, how employers interpret them, and why similar capabilities can lead to very different employment outcomes.

The deeper point is that employability policy needs to look beyond skill acquisition and ask whether capability survives, remains relevant, is recognised and can actually be converted into opportunity.

A Skill Can Be Relevant at Graduation and Less Useful Months Later

Most debates about graduate employability begin with a mismatch question: are universities teaching what employers need? The DDSRF treats that as only the first stage. A strong match at one moment can weaken later because both the graduate and the labour market continue changing.

The study separates two mechanisms behind that movement. Digital skill decay refers to a loss of retained proficiency or usability, including through non-use or insufficient practice. Digital skill obsolescence refers to changes in technologies, tasks or standards that reduce the value of a capability even when the graduate still possesses it. This means a university can teach a relevant skill well and still see graduates struggle later. The problem may not be poor teaching at all. A capability can remain technically intact while becoming less useful because workplace systems, tools or task requirements have moved on.

Specialised digital capabilities may face particular exposure because their usefulness can depend heavily on a specific technological environment. The same specialisation that improves initial job fit can become a vulnerability when tools or work processes change quickly.

Graduate behaviour also affects what happens next. Skills can be refreshed, recombined or supplemented before labour-market entry. Changes in demand can even raise the value of capabilities that were previously peripheral. Employability thus depends on how capability and demand evolve together, not simply on what was taught at the point of graduation.

Employers Judge Evidence, Not Invisible Capability

Graduates do not enter recruitment as transparent bundles of skills. Employers cannot directly observe everything a candidate knows, so they rely on degrees, work experience, portfolios, projects, micro-credentials and other signals that help them infer competence and relevance.

The DDSRF gives employer judgement a central place in the pathway from education to employment. A graduate first needs to make a capability visible. Employers then decide whether the evidence is credible and whether the capability is sufficiently relevant to the role they are filling.

Recognition and value are not the same thing. An employer may believe that a graduate possesses a particular capability but still attach limited importance to it in a given recruitment setting. Skill possession can therefore remain economically weak even when it is real and demonstrable.

For universities, that changes the role of employability support. Work-integrated learning, relevant placements, portfolios and micro-credentials are useful only when they help graduates present capability in forms that employers can understand and trust.

Credential expansion on its own offers no guarantee. Producing more certificates may increase the amount of evidence a graduate carries without improving employer confidence or job relevance. The more useful question is whether the evidence changes how a capability is interpreted during recruitment.

Equal Capability Does Not Produce Equal Opportunity

A skills-only view also struggles to explain why graduates with similar capabilities can experience very different transitions into work. Institutional background, field of study, socioeconomic resources, access to networks and local labour-market conditions can all influence whether capability is converted into a worthwhile employment outcome.

Institutional prestige and networks may affect access to opportunities before specific competencies are closely examined. Differences between faculties can influence work readiness and recruitment exposure, while socioeconomic resources can shape who obtains placements, professional contacts and other experiences that strengthen labour-market signals.

Local conditions add another layer. The same capability can carry different value depending on the availability of suitable roles, the structure of regional demand and the recruitment environment in which graduates compete.

The DDSRF treats employability as more than whether someone gets a job. It separates labour-market access, capability-to-role fit, job and transition quality, and early-career progression. A graduate can succeed on one measure while performing poorly on another. This broader view has major policy implications. Weak graduate outcomes should not automatically trigger calls for more training. The problem may sit in employer recognition, local demand, access to opportunity or the conditions surrounding recruitment rather than in the capability itself.

For developing countries and institutions operating in uneven labour markets, the framework is especially relevant. Expanding digital education may improve capability formation, but employment gains can remain limited where graduates lack access to strong signalling opportunities or where suitable demand is concentrated in only part of the economy.

Universities Need Better Diagnosis Before They Rewrite the Curriculum

Poor graduate outcomes can emerge from very different failure points. Weak capability formation calls for changes in teaching or curriculum design. Declining relevance may require skill refresh and continuing learning. Poor visibility points toward better ways of demonstrating capability, while unequal outcomes may require action on access, networks and opportunity.

Universities thus need evidence that shows where value is being lost. Graduate-outcome tracking, employer engagement, work-integrated-learning partnerships, programme review and labour-market information can help identify recurring problems across cohorts rather than relying on isolated complaints.

Employer feedback also needs careful handling. A single company request may reflect a temporary shortage, a specific technology stack or a recruitment preference that is unlikely to persist. Repeated signals across employers and graduate outcomes provide a stronger basis for programme change.

The study cautions against turning degree programmes into direct reflections of current vacancies. Higher education operates on longer cycles than many digital markets, and constant redesign risks making curricula more reactive without making graduates more resilient.

Transferable digital foundations, opportunities for updating, continuing-learning routes and better mechanisms for demonstrating capability may offer more durable protection against labour-market volatility. Universities can improve relevance without making short-term employer demand the organising principle of an entire degree.

The framework itself remains untested as a complete causal model. It is a conceptual synthesis built from targeted theory-building searches rather than an exhaustive systematic review, and the evidence base does not yet provide strong longitudinal tracking from skill acquisition through labour-market entry.

Testing its claims will require repeated measurement of both graduate capability and changing occupational demand. Researchers will also need to examine whether identical capabilities receive different employer responses depending on how they are signalled and whether comparable recognised skills produce different outcomes across institutions, socioeconomic groups and labour markets.

The most important implication is a change in what counts as graduate readiness. Producing more digital skills is not enough when those skills can weaken, lose relevance, remain hidden from employers or fail to generate opportunity under unequal conditions.