Labour scarcity in the Netherlands is no longer a phase of the business cycle. Supply barely grows before 2040, so demand adapts: firms turn work down, raise their prices or automate. Our conclusion: the question for employers shifts from how to fill a vacancy to which work can still be done, and the people who could do it are mostly not looking.
The turning point
In late 2021, for the first time since Dutch measurement began, open vacancies outnumbered unemployed people. The central bank DNB showed in 2024 that this is structural. Ageing and the low birth rates of earlier decades bring growth in labour supply close to a standstill after 2030. Raising participation offers little room, because Dutch participation is already among the highest in the world.
DNB ran the arithmetic on the ways out. Far higher participation, much longer working weeks, a large new wave of labour migration or a sharp rise in productivity growth would each be needed to keep income per head growing. Each route runs into a limit: people who are not working are mostly retired, ill or studying; part-timers want only a little more work; the countries that supply labour migrants are ageing faster than the Netherlands; and productivity growth would have to break a long trend.
How demand adapts
Business services shows what adaptation looks like, because there the staff are the product. ING found that a large share of firms in the sector were short of people at the end of 2025, from cleaning and security to accountancy and law.
The result is not only that vacancies stay open longer. Firms take on less work: they accept fewer new clients and choose their assignments more carefully. Growth is now limited by whether the work can be staffed. Prices rise too, with cleaning, accountancy and legal rates all moving up. And automation splits the sector: AI can absorb routine work for accountants and lawyers, while cleaning and security stay dependent on people.
A problem of fit, not of numbers
In pure numbers, the gap between unemployed people and open vacancies is small. In skills it is wide. Eurostat measures a broader labour market slack, including part-timers who want more hours and people who are available but not looking, and that slack is far larger than recorded unemployment. The people exist. They are simply not visible as job seekers.
Mobility helps only partly. People change employer often, but mostly within related sectors, rarely from a shrinking sector to a growing one. Training is the hinge, and it is unevenly spread: people on flexible contracts and new starters learn more, while those who stay longest in one place develop least.
What this does not show
The structural analysis is a DNB scenario from 2024, built on reference periods between 2021 and 2023. It assumes current trends in migration, participation and productivity. A strong productivity gain from AI, or a sustained rise in migration, would change it. The business services figures come from a single sector study by ING and cannot be stretched to the whole economy. Our reading weakens if labour supply turns out to grow more than DNB expects.
What you can do
- Count the work you turn away as a cost. When an open role means a lost client, the real cost of a vacancy is the revenue you miss.
- Ask your part-timers whether they want more hours. Across a team, a modest increase per person can equal a new hire.
- Approach people who are not looking. The reserve sits outside job boards, among people in work who would move for the right role.
- Hire across sector lines and train for the gap. A candidate from an adjacent sector with a learning plan beats a vacancy that stays open.
- Invest in training for long-serving staff, the group that develops least on its own.