# Graduates Without Jobs, Part 3: A Five-Layer Plan to Match Degrees with Jobs

URL: https://edufabtech.com/blog/graduates-without-jobs-part-3-five-layer-plan  
Section: Blog · Education Systems  
Published: 2026-11-13  
Publisher: EduFabTech (https://edufabtech.com)

> The final part of our series on graduate unemployment: a five-layer plan built on ISCO, ISCED, ESCO and the UNESCO Global Convention, tested against Germany, Switzerland, Singapore, Finland, South Korea and Türkiye, with to-do lists for students, universities, companies and governments.

Every year, Germany runs two counts that should not exist side by side. Firms report training places they cannot fill, and young people report that they cannot find a training place. In 2025, about 84,400 young people ended their search for an apprenticeship without success, the Federal Institute for Vocational Education and Training (BIBB) reported, while the supply of places and the number of new contracts both fell again. Jobs and people were both there. They did not meet.

That mismatch, repeated across countries and across levels of education, is what this series has been about. [Part 1](https://edufabtech.com/blog/graduates-without-jobs-part-1-264-million-students) measured the problem: 264 million students in higher education, graduate unemployment above that of less-educated young people in most regions of the world, and about half of new US graduates working in jobs that do not need a degree. [Part 2](https://edufabtech.com/blog/graduates-without-jobs-part-2-crossing-fields) followed the graduates who cross into other fields and set out a 90-day plan for making that move on purpose. This final part asks what countries can do so that fewer graduates need a rescue plan at all.

The answer proposed here is a plan in five layers. It starts with a common language, because a country cannot plan what it cannot name, and it ends with international mobility, because talent no longer stays inside borders. For each layer, the plan says who acts, how long it takes, where it has been tried and what happened. No country gets every layer right, and the weak points are part of the evidence.

## Layer 1: a common language for jobs, skills and qualifications

Employers describe jobs, universities describe programmes, and statistics offices describe both, often in three different vocabularies. International classifications already exist to close that gap. The International Labour Organization's ISCO-08, adopted in December 2007 and endorsed in March 2008, sorts all work into 10 major groups and 436 unit groups. UNESCO's ISCED 2011 does the same for levels of education, from early childhood to doctorate. Below them sit skills taxonomies: the EU's ESCO, whose skills pillar holds 13,485 skill and knowledge concepts linked to occupations, and the US O\*NET, which publishes detailed data for 912 occupations in its August 2026 release.

The third piece is recognition. UNESCO's Global Convention on the Recognition of Qualifications concerning Higher Education, the first UN treaty on higher education, entered into force in March 2023. By early 2026 it had 40 states parties, hosting about 2.2 million of the world's 7.3 million internationally mobile students, UNESCO says.

**Who:** the national statistics office, the education ministry and the body that awards qualifications. **How long:** two to three years to map national job titles to ISCO and degree programmes to ISCED and to a skills taxonomy; ratifying the convention runs in parallel. **Tried in:** the EU, the United States, and the convention's states parties. **Result and weak point:** a shared vocabulary makes job adverts, degrees and statistics comparable, but it does not create jobs, and recognition on paper is not recognition in practice. Germany, which has a federal law on recognising foreign qualifications, processed 55,332 applications in 2024; 43.2 per cent ended in full equivalence and 46.6 per cent in a demand for further training or tests, the BQ-Portal reports. Doctors trained abroad have described waits of two years or more to The Local.

## Layer 2: an annual skills needs report

Once jobs and skills share one language, a country can count them every year. The EU agency Cedefop already publishes forecasts for every member state. Its 2025 edition projects marginal EU employment growth of 0.4 per cent a year to 2035, and finds that shortages in health, care and construction come mainly from workers retiring, not from new jobs. That is an important point for planners: most openings replace people, so the question is less "which new industry?" than "who replaces the nurses and electricians who leave?". The OECD's Skills Strategy, applied in 11 countries between 2012 and 2019, and the ILO's skills anticipation guides offer the method for countries that start from scratch.

**Who:** an independent labour market observatory, with employers, unions and universities at the table. **How long:** a first report within 12 to 18 months; three to five years before the time series is reliable. **Tried in:** all EU member states through Cedefop. **Weak point:** forecasts are built on past trends and can miss shocks, such as the arrival of generative AI in entry-level work. A skills report should guide decisions, not set fixed quotas.

![Comparison of six country models, each with one figure and one weak point: Germany, Switzerland, Singapore, Finland, South Korea and Türkiye](https://edufabtech.com/uploads/2026/11/graduates-without-jobs-part-3-five-layer-plan-2.png)
*No country gets every layer right: six models, each with one headline figure and one weak point. — EduFabTech · Own work*

## Layer 3: education that is built with employers

This is the layer with the most evidence and the most cost. In Switzerland, around two-thirds of young people enter vocational education and training after compulsory school, choosing among about 250 occupations, the State Secretariat for Education, Research and Innovation (SERI) reports. Most host firms recover more than they spend, because apprentices produce more than their training costs, and the system stays open: holders of a vocational baccalaureate can enter a university of applied sciences without an exam.

In Germany the books are less favourable to firms. BIBB's cost-benefit survey for 2022/23 puts the gross cost of an apprentice at €26,210 a year, of which €18,124, or 69 per cent, is recovered through the apprentice's work, leaving a net cost of €8,086. A year after finishing, 62 per cent of trainees were still with the firm that trained them. Germany had the lowest youth unemployment in the EU in July 2026, at 7.2 per cent, against an EU average of 15.1 per cent and 22.9 per cent in Spain, Eurostat reports.

Neither model is ideal. Both ask children to choose a direction at around 15, and Germany's training market is shrinking, which is why 84,400 young people were left out in 2025. Other countries show what happens when one piece is copied without the rest. South Korea has the highest share of young adults with a tertiary degree in the OECD, 71.1 per cent of 25- to 34-year-olds against an OECD average of 47.5 per cent, but the employment rate of its university graduates aged 25 to 64 is 80.1 per cent, below the OECD's 86.0 per cent, Herald Business reported from Education at a Glance 2026. Korea's employer-linked Meister high schools placed 71.8 per cent of their 2026 graduates in jobs, the lowest rate in five years. In Türkiye, 392,887 students aged 15 to 18 attended vocational education centres (MESEM) in 2024–25, spending four days a week at a workplace, and the workers' rights group İSİG recorded at least 15 MESEM students killed at work over two school years, bianet reported. Work-based learning without inspection and fair pay becomes cheap labour. And in England, an apprenticeship levy on large employers was followed by a fall in starts of about a third between 2016/17, when the levy began, and 2020/21, a period that included the pandemic; in 2024/25, 39.8 per cent of starts were at degree-equivalent levels and only 21.2 per cent went to under-19s, official data show.

**Who:** the education ministry, universities, and chambers of commerce or industry bodies. **How long:** five to ten years for a dual system to mature. **What to do:** tie part of university funding to published graduate outcomes rather than ordering quotas; build a paid, inspected work placement into every degree; and let employers co-design courses where the skills report shows a gap.

## Layer 4: a bridge between graduation and the first job

Even a well-planned system leaves gaps, so the fourth layer catches graduates in the months after graduation. Singapore's SkillsFuture, launched in 2015, gives every adult a training credit. In 2025, 606,000 people took part in subsidised training, and of 15,000 people who completed its Career Transition Programme between June 2022 and June 2025, 51 per cent found new roles within six months, SkillsFuture Singapore reports. The weak point came early: in 2017 a criminal syndicate used forged documents to obtain close to S$40 million in training subsidies, SSG disclosed, and the agency had to rebuild its checks.

Finland introduced a youth guarantee in 2013, promising every young person a job, training or a workshop place. The National Audit Office found that its targets were not met in its first year and that its effect on the length of unemployment was roughly zero, largely because employment offices were cutting staff while the guarantee was being launched. Employment plans were drawn up for 92.6 per cent of young jobseekers within three months, but the auditors described them as brief and cursory. Finland's youth unemployment stood at 23.4 per cent in July 2026, among the highest in the EU. A guarantee is only as good as the staff and the jobs behind it.

**Who:** public employment services and employers. **How long:** one to two years to set up; each cohort runs six to twelve months. **What works:** short, paid placements tied to real vacancies, funding paid on results, and fraud checks from day one.

![Timeline over ten years for the five layers: common language in years 0 to 3, first skills report by year 1.5, dual training maturing from year 5, bridge programmes set up in two years, a two-year brain-circulation pilot from year 2](https://edufabtech.com/uploads/2026/11/graduates-without-jobs-part-3-five-layer-plan-3.png)
*An implementation timeline for the five layers. Durations are EduFabTech estimates drawn from the evidence in the article. — EduFabTech · Own work*

## Layer 5: brain circulation instead of brain drain

Countries that train more graduates than their economies can absorb lose them abroad, and countries that are ageing need them. The final layer turns that one-way flow into a partnership. In the PALIM pilot between Belgium and Morocco, run from 2019 to 2020, 120 young Moroccans were selected from 9,677 applicants and trained in software development, Dutch and workplace culture, with some placed in Flanders and the rest in Morocco, the Center for Global Development records. The idea, known as a Global Skill Partnership, is that the destination country helps pay for training in the origin country, so both gain skilled workers.

**Who:** governments of origin and destination, employers and training providers. **How long:** a two-year pilot, then scale-up. **Weak point:** the pilots are tiny, and PALIM's placements were disrupted by the COVID-19 pandemic. Without layer 1, a partnership like this stalls at the recognition desk.

## Is this realistic?

Four obstacles stand in the way, and each has a known mitigation.

- **Cost.** A German apprentice costs a firm €8,086 a year net. Mitigation: share the cost through training funds paid by all firms in a sector, so that firms that train are not subsidising those that only poach.
- **Political cycles.** A dual system takes longer to build than most governments last. Mitigation: put the skills report and the common language in law, run by an independent body with employers and unions on its board.
- **Missing data.** In 2013, Finland's statistics office counted about 65,000 unemployed under-25s while the employment ministry counted about 39,000, because they defined students differently. Mitigation: layer 1 first; one definition, one count.
- **University autonomy.** Orders to cut or expand programmes meet resistance, and rightly so, because forecasts can be wrong. Mitigation: publish graduate outcomes for every programme and link part of the funding to them, leaving the decisions to universities.

The risks are just as concrete: forecasts that miss the next shock, training that becomes cheap labour, subsidies that attract fraud, and pressure on teenagers to choose too early. Each of them has already happened somewhere in this article. That is the strongest argument for building the plan in layers with evaluation at every step, rather than copying one country's model whole.

## Who does what

### Students

- Read your country's skills report, or Cedefop's for your country, before choosing a programme.
- Learn the ISCO and ESCO names for the jobs you want; that is the language employers' systems use.
- Do at least one paid, supervised work placement before you graduate.
- If you plan to work abroad, check whether your country has ratified the Global Convention.

### Universities

- Publish employment outcomes for every programme, one and five years after graduation.
- Map each course to skills in ESCO or O\*NET, so employers can read your degrees.
- Build placements into the timetable, with a named employer partner for each programme.

### Companies

- Write job adverts in standard skill terms, not in-house jargon.
- Take apprentices and interns as an investment, with pay, a mentor and safety training.
- Sit on the board that writes the skills report, and share vacancy data with it.

### Governments

- Adopt ISCO-08 and ISCED 2011 across all ministries, and ratify the Global Convention.
- Fund an independent observatory to publish a skills report every year.
- Inspect every workplace that hosts students, and evaluate every programme against its targets.

## The end of the series, and the start of a plan

Graduate unemployment is not a verdict on young people or on degrees. It is the result of education systems and labour markets that grew apart and stopped speaking the same language. The tools to bring them back together already exist: international classifications, annual forecasts, work-based learning, bridge programmes and skill partnerships. Every one of them has worked somewhere, and every one has failed somewhere. A country that builds all five layers, measures them and fixes what breaks will waste less of its young people's time.

Which layer is missing in your country? Tell us in the [forum](https://edufabtech.com/forum), and answer this week's question in our [polls](https://edufabtech.com/polls): should universities be funded partly on their graduates' employment rates?

## Sources

1. UNESCO. (2026). Global Convention on the Recognition of Qualifications concerning Higher Education: keynote to INQAAHE. UNESCO; INQAAHE https://www.inqaahe.org/wp-content/uploads/2026/02/INQAAHE-Keynote-by-UNESCO-5-Feb-2026.pdf
2. Federal Institute for Vocational Education and Training (BIBB). (2025). Apprenticeship training or external recruitment of skilled workers: what costs should firms expect? BIBB Cost-Benefit Survey 2022/2023. BIBB https://www.bibb.de/dienst/publikationen/en/21068
3. Eurostat. (2026). Euro area unemployment, July 2026. Eurostat https://ec.europa.eu/eurostat/web/products-euro-indicators/w/3-01092026-bp
4. SkillsFuture Singapore. (2026). More than 1 in 2 eligible Singaporeans aged 30-75 utilised SkillsFuture Credit as the SkillsFuture movement marked its 10th anniversary. SkillsFuture Singapore https://www.ssg.gov.sg/newsroom/more-than-1-in-2-eligible-singaporeans-aged-30-75-utilised-skillsfuture-credit-as-the-skillsfuture-movement-marked-its-10th-anniversary/
5. Cedefop. (2025). Cedefop skills forecast 2035: twin transition and demographic challenge drive demand for high-level skills. Cedefop https://www.cedefop.europa.eu/mt/news/cedefop-skills-forecast-2035-twin-transition-and-demographic-challenge-drive-demand-high-level

---
Source page: https://edufabtech.com/blog/graduates-without-jobs-part-3-five-layer-plan  
Quote with a link to the source page. © EduFabTech.
