36 publications
Every report and article from IPMERC Research in one place. Public, sourced, and free to use.
In 2025 the Netherlands counted 372,000 part-timers who want to work more hours, against 297,000 unemployed people. The largest untapped group on the labour market already has a job. Part-timers say they want 2 to 2.5 extra hours a week on average. At a client with twenty part-timers that is more than a full working week, without a single new candidate.
Construction needs 75,000 new full-time workers between 2026 and 2029, of which education can supply roughly 50,000 and 25,000 must come from lateral entry. Grid operators need 28,000 technicians by 2029, 23,000 of them at contractors. Healthcare projects a shortfall of 155,000 by 2032, against 37,000 in 2021. These are the three sectors where the vacancy rate barely fell after 2022.
Between mid-2022 and mid-2025 the Dutch vacancy rate fell from 5.1% to 4.2%. That average hides the real story. In ICT the rate dropped from 8.2% to 5.1%, in trade and hospitality from 6.6% to 4.7%. In construction it went from 7.5% to 7.4%, and in healthcare from 4.4% to 4.3%. Two sectors barely cooled at all, and they are exactly the sectors where production cannot be postponed or automated.
The Dutch IT labour market has been cooling since 2023. Yet employment in IT occupations grew by 36,000 year on year in the second quarter of 2026, the largest increase of any occupational group. Those figures do not contradict each other: the market is normalising after an exceptional peak, but the constraint has moved from volume to fit. In 2024, 63% of companies that tried to recruit IT staff failed to fill every vacancy.
In the fourth quarter of 2024, 34 of the 93 scored Dutch occupation groups were tight and 56 very tight; 2 were average and 1 was loose. For 32 of the 112 groups, ROA forecasts further tightening to 2030. What sets those groups apart is rarely job growth and usually departure: on average 2.8% of workers must be replaced each year, among butchers 8.1%, the highest of all occupation groups, and among advisers in marketing, public relations, and sales 1.1%. Butchers see almost no job growth and stay tight anyway; the advisers ease slightly. Representatives and buyers shrink by 1,900 workers and still stay on the tight side. The three scored ICT groups, together 518,400 workers, were all very tight at the end of 2024 and all ease slightly to 2030; for ICT user support that was already visible in the third quarter of 2025. For five technical occupation groups, the forecast easing was not yet visible in the Spanningsindicator in that same quarter.
In 2024, 71.0% of Dutch enterprises with 10 or more employees bought paid cloud services. In the same year 29.7% of them employed an ICT specialist. By 2025, 55.4% of Dutch enterprises hosted a database or their files with a cloud provider against 24.0% across the European Union, and only Denmark was higher at 55.5%. Migration off those services requires ICT staff. Of the Dutch enterprises that tried to recruit an ICT specialist in 2024, 63.1% were left with a vacancy they could not fill, and the reason cited most often was that nobody applied. When the Netherlands Authority for Consumers and Markets surveyed 420 business cloud users, 172 had at some point tried to switch provider. For 52 of them the attempt failed, which is 30.2% of all attempts. Among those that did switch, 52.5% used an intermediary.
Labour scarcity is no longer a matter of the business cycle. The Dutch central bank concluded in 2024 that an ageing population leaves labour supply barely growing through to 2040. If supply stops growing, demand adapts: firms scale work back, raise its price, or automate it. Business services already shows this. In numbers the gap is nearly closed; in quality it is wide.
When Dutch employers were asked in autumn 2025 to name their hardest vacancy, 27% named a technical occupation: fitters, welders, CNC operators, machinists, and engineers. Care and welfare professions followed at 13%. In construction, 71% of vacancies arising over the past 12 months were hard to fill; in manufacturing, 53%. The vacancy rate in the fourth quarter of 2025 stood at 7.0% in construction, 5.0% in professional and technical services, and 4.8% in ICT, against 3.9% for the economy as a whole. The overall market is loosening, the share of hard-to-fill vacancies fell from 53% in 2023 to 45% in 2025, but that is largely an office-job story. In Germany, the ICT vacancy rate dropped to 2.5%. The Dutch technical shortage is not a European inevitability. It is a Dutch profile.
Of the 9.42 million Dutch people working between the ages of 15 and 64, 1.87 million are between 55 and 64: 19.9%. They leave the labour market within roughly ten years. That is not a cycle and not a growth scenario, it is an age structure already fixed. The demand for staff it creates exists regardless of what the economy does, while the inflow meant to absorb it is smaller than the outflow.
The Netherlands produced 8,710 ICT graduates at bachelor and master level in 2024, against 3,101 in 2015. That is a rise of 181% in nine years, far faster than the 28% growth in total graduates. Yet the share of companies with hard-to-fill IT vacancies stayed around 63%. More supply from education has not resolved the shortage, and this report sets out why.
The Dutch labour cost index went from 113.1 to 143.1 between mid-2021 and mid-2025, a rise of 26.5%. That is more than Germany at 23.1% and nearly double France at 14.1%. Yet only 39% of Dutch employers name candidate salary expectations as a reason an IT vacancy is hard to fill, below the EU average of 42%. Together those facts mean paying more does not fix a matching problem.
How you ask for a starting salary decides whether asking pays. In a study of 149 American professionals, published in the Journal of Organizational Behavior in 2011, those who negotiated gained an average of $5,000 over the first offer. The gains came from two of the five styles measured: competing and collaborating. Compromising and accommodating produced no measurable gain, and risk-averse participants negotiated least and were least satisfied afterwards. A $5,000 difference at the start compounds to roughly $604,000 over forty years at 5% annual raises, by our own arithmetic.
Every quarter, hundreds of thousands of Dutch employees start work for a new employer. In the second quarter of 2026 the count was 320,000, or 4.0% of all employees, roughly one in 25. Since 2013 the quarterly rate has moved between 2.6% and 5.3%, and the tight-market peak has fully unwound: the current pace sits almost exactly on the 2019 level. The flow is carried by the flexible layer: more than six in ten switchers leave a flexible contract, and 85% start on one again after the move.
Of Dutch employers with external vacancies, 65% made extra recruitment efforts in autumn 2025 because of labour market tightness. The order forms a ladder. At the top sit the cheap rungs: spreading vacancies across more channels (54%), recruiting through the employer's own network (52%), and hiring candidates who still need training (52%). Better terms of employment follow at 34%. Only then comes the outside door: 22% hired more temporary, seconded, or freelance staff, and 21% deployed a recruiter, recruitment agency, or headhunter more often. In manufacturing, both external routes sit above average. At the peak of the shortage, in autumn 2021, half of employers with hard-to-fill vacancies engaged a temp agency, recruitment agency, or headhunter. An agency that wins a mandate therefore receives, on average, a vacancy the market has already rejected: 87% of employers with recruitment problems reported too few applications.
A Dutch recruitment agency typically charges 22 to 32% of gross annual salary per placement on a contingency basis; retained search runs at 20 to 30%, with part of the fee paid up front. On a €55,000 salary at a 25% fee that is €13,750. An in-house corporate recruiter costs €65,000 to €80,000 a year, built from an average salary of roughly €41,000 to €55,000 (consulted August 2026), about 30% in employer charges, and several thousand euros in systems and job boards. That fixed cost only pays for itself from about five to seven placements a year that would otherwise have gone through an agency. Above that volume the in-house recruiter wins comfortably: the most recent Dutch benchmark (measured in 2020) put corporate cost per hire at €3,818, with recruiters carrying an average portfolio of 20 vacancies. Below that volume, a company pays more for its own recruiter than it would have paid in fees. And for scarce technical profiles the most expensive outcome looms: months of internal effort first, then the agency fee anyway.
In 2026, platform vendor Phenom audited 219 employers across eight industries on the automation of their hiring process. On attracting, engaging, and converting candidates, employers average 62% of the attainable maximum. On qualification after the apply click, the same employers average 21%, a gap of 41 percentage points. 94% do not schedule an interview at the point of application. 99% do not use video interviews inline. Less than 1% have a fully orchestrated qualification process. For frontline roles, 85% sit in the lowest automation tier. For knowledge worker roles the figure is 93%. No industry scores above 30%. The data comes from a platform vendor measuring against its own product categories, and that caveat applies to every use of the figures.
A 2026 survey of more than a hundred organisations by Aptitude Research shows that 35% of recruiter time goes to interview coordination, 25% to screening, and 24% to candidate communication. Of the 219 employers Phenom audited, 94% do not offer automated interview scheduling at the point of application. 6% do. 1% deploy a voice screening agent. 57% of surveyed organisations already use some form of automation agent, but adoption clusters at the front of the process, not at the coordination steps that consume the largest share of hours.
A 2026 survey of more than a hundred organisations by Aptitude Research shows that 54% name quality as the biggest hiring challenge. 45% name speed and 39% name cost. Among employers that rate their automation as effective, 42% report higher quality of hire. But 61% of the 219 audited employers apply the same automation to frontline and knowledge worker roles, with no distinction by profile type. 64% measure automation usage well, but only 60% do so consistently across role types. The shift from speed to quality changes what a client expects from an agency.
The Netherlands is Europe's home-working country. In 2025, 11.3% of employed people aged 15 to 64 usually worked from home and 41.0% sometimes did, together 52.3%. Sweden follows at a distance with 44.5% and the EU average stands at 23.0%. The total has been nearly stable since 2021, but the mix shifted: fully remote fell from 20.4% to 11.3%, while occasional home working rose from 33.5% to 41.0%. Hybrid won. For recruitment this means commuting distance matters less in the Netherlands than anywhere else in Europe.
Unemployment in the Netherlands is mostly a pass-through. Over the four quarters to the first quarter of 2026, an average of 37.0% of unemployed people were in work one quarter later, more than one in three. The EU average stands at 23.1%, and only Denmark sits higher at 39.2%. At the peak of the tight market the Dutch outflow was higher still: 41.5% over the same quarters in 2022 and 2023. The other side is small but growing: the share of workers becoming unemployed rose within a year from 1.0% to 1.4% per quarter.
Europe officially counts 2,617 reported shortage occupations and 2,177 surplus occupations, and the same occupations regularly appear on a shortage list in one country and a surplus list in another. The Netherlands is an extreme: 195 shortage occupations, second only to Italy, against 8 surplus occupations, nearly the fewest in Europe. There is almost nothing left to reallocate domestically. Around 53 million European workers, about a quarter of employment, sit in an occupation with widespread shortages or surpluses.
Nowhere in the EU does working life last as long as in the Netherlands: 44.0 expected years in 2025, against 43.4 in Sweden and 37.5 on average in the EU. More striking is the pace at which it happened. In 2000 the Netherlands stood at 35.5 years; 8.5 years were added in 25 years, driven by women's labour participation and later retirement. For replacement demand this is the softest cushion there is: people keep working longer. But the cushion is largely used up, because a working life can hardly be stretched much further.
The Dutch economy runs on technology, but its working population is thinly educated in it. In 2025, 9.0% of employed people held a completed tertiary degree in the STEM fields of science, engineering or ICT, against 11.1% on average in the EU. That puts the Netherlands 22nd of the 30 measured countries; Lithuania leads with 15.1%. The stock is growing, from 748,700 workers in 2021 to 869,500 in 2025, a rise of 16%. Of them, 65% work as professionals and some 12% outside the knowledge occupations.
Every plan that counts on activating young people runs, in the Netherlands, into an empty reserve. Of people aged 15 to 29, 5.3% were neither in work nor in education or training in 2025, the lowest share in the EU; the average is 11.0% and Sweden follows at 5.9%. Within that small group, most are willing: 4.3 of the 5.3 percentage points say they want to work. The reserve is not unwilling but thin, and whoever recruits from it competes with all of Europe for the continent's smallest surplus.
The European outlook for the Dutch labour market to 2035 confirms what the replacement figures already showed, and extends it a decade. Of all expected job openings between 2022 and 2035, 68% stem from replacing departing workers; among technicians and associate professionals it is 91%. Professionals account for 44% of all openings and 72% of openings require high qualifications, some 12 points above the EU average. The Dutch labour force aged 55 and over grows 16% to 2035, against just under 10% in the EU.
Every quarter the European Commission asks firms what limits their production. In July 2026 three of the four constraints in industry fell: insufficient demand to 33.9%, the lowest since July 2023, material shortages to 12.6% and financial constraints to 5.2%. Only the shortage of labour rose, by 0.6 points to 17.1%. In construction, 28.5% of firms called the labour shortage a constraint in December 2025, nearly as many as the 31.0% naming insufficient demand. The economy is picking up, and the first thing that pinches then is staff.
Nowhere in the EU do as many people stack jobs as in the Netherlands. In 2025, 983,000 workers aged 15 to 64 held a second job, 10.4% of all employed people, against an EU average of 4.1%. Denmark follows at 9.7%, Germany stands at 5.1%. The explanation sits close to the part-time structure: where main jobs are small, there is room to stack. For recruitment the second job is an underrated signal: nearly a million people demonstrate they can and will work more hours than their main job offers.
While the IT vacancy rate fell sharply after 2022, the business register kept moving the other way: from 92,620 ICT enterprises in 2021 to 111,713 in 2024, growth of 21% in three years. Employment in the sector grew along to 386,306 people in 2023, 8% more than in 2021, at fewer than four people per enterprise. The sector stands at 3.8% of Dutch employment, against 3.4% on average in the EU, and 5.3% of value added. The growth sits mostly in very small firms, and that is the real finding.
While 77% of European employers already struggled in 2019 to find people with the right skills, good documentation of what firms actually do about it is scarce. The European foundation for working conditions studied it at 17 organisations across 13 member states. The measures cluster in four groups: partnering with education and intermediaries, offering more than pay, recruiting smarter through referrals and wider catchment areas, and selecting on aptitude rather than diplomas. The most striking finding: the will to recruit internationally exceeds its use.
The catch-up in Dutch collectively agreed wages is past its top. After growth of 6.0% in 2023 and 6.6% in 2024, the crest of the wave, the annual rate sank to 5.0% in 2025 and 3.9% in July 2026. The euro area moves the same way: the ECB wage tracker, fed by collective agreements from nine countries including the Netherlands, points to 2.3% for 2026 and 2.7% for early 2027, against 3.2% in 2025. Dutch wages still grow faster than the currency area's, but the gap narrows and the direction is down everywhere.
Whoever reads Dutch scarcity as a local problem misses half the story. Across the OECD, unemployment stood at 4.9% in May 2026, close to its lowest point since records began, while employment and participation reached first-quarter records of 72.1% and 76.7%. The Netherlands sits below that at 3.8% unemployment in June, against 6.0% in the EU. At the same time employment growth slows, with 0.3% expected for 2026, and real wages in a third of member countries still sat below their early 2021 level.
In the fourth quarter of 2025 the Netherlands had 377,000 job vacancies against 407,000 unemployed people: 93 vacancies per 100 unemployed, the highest ratio of the 24 member states that publish both series. Malta follows on 80 and Germany on 66; Romania closes the list on 6, or eighteen unemployed people per vacancy. At the European peak in mid-2022 the Netherlands ranked third, behind Czechia (186) and Germany (140). Since then the Dutch ratio has fallen by 31%, against 53% in Germany, 60% in Austria and 72% in Czechia. Elsewhere that fall tracks economic growth (correlation 0.70 across 24 countries), yet the Dutch economy grew only 2.1% over the same period. What sets the Netherlands apart is the level it settles at: a job vacancy rate of 4.0% against its own pre-pandemic record of 3.3%, on the highest part-time share in the EU at 42.3%.
In 2025, 78.7% of Dutch employees in ICT occupations say they are satisfied with their work, against 78.6% of all employees. In 2018 IT led by 4.4 points. That lead is gone while the lead in freedom and pay stayed: 79.8% of IT workers regularly decide how they do their work, against 60.2% of all employees, and the median hourly wage is €34.80 against €26.90. Burn-out complaints run ahead instead, at 24.1% against 20.7%. This paper sets the figures of the Dutch working conditions survey beside the model of labour economist Peter van der Meer, which explains why pay and autonomy do not carry wellbeing at work on their own, and what that means for anyone recruiting IT staff.
Seven Dutch regulators that have kept the same form since 2010 grew from 2,730 to 4,772 full-time staff, a rise of 75%. The data protection authority went from 79 to 312, the mines inspectorate from 57 to 188, the healthcare authority from 254 to 521. The financial markets regulator AFM grew 70% and the central bank DNB 54%. The central government as a whole employed 160,016 full-time staff at the end of 2025 against 114,328 at the end of 2010, up 40.0%. The whole Dutch economy grew 19.6% in full-time years over the same period. Inside the central government, inspection work grew 33.5% between 2021 and 2025, against 22.0% for all staff. The growth tracks new legal duties: the GDPR in 2018, online gambling in 2021, the Groningen earthquakes, and new European financial rules. The Jetten cabinet books €1.4bn of savings on the government apparatus in 2030. The CPB, the government's own forecaster, counts €0.2bn of it and calls the rest implausible while no tasks are dropped. Growth already stalled in 2025: the competition authority shrank, the health inspectorate shrank, the healthcare authority stood still, and the central government added 1.9%.
In November 2025 the Netherlands counted 55.6 benefit recipients for every 100 people whose main income is work: 39.6 on a pension and 16.0 on a social security benefit. In 2014 the figure was 63.6. Since 2001 the country gained 940,000 pensioners and their share of the population rose from 15.1% to 18.6%, but the number of workers grew faster, from 7.14 million in 2014 to 8.51 million in 2025. That growth is largely older people. Of those aged 60 to 65, 36.9% worked in 2010 and 69.5% in 2025. Of those aged 65 to 70 the share went from 12.1% to 29.5%. The share of 55 to 65 year olds living mainly on a pension fell from 18.9% in 2006 to 2.9% in 2025, while the state pension age rose from 65 to 67. The cabinet counted on still more supply from tying the pension age one-to-one to life expectancy, then dropped that plan on 26 May 2026. What remains, a stricter entry test for unemployment benefit, would exclude about 12% of new claimants according to UWV, half of them on temporary contracts.
The Netherlands employed 215,000 administrative clerks in 2025 against 283,000 in 2013, a fall of 24.0% over a period in which the employed labour force grew 16.6%. Secretarial staff went from 64,000 to 35,000, down 45.3%. Administrative personnel as a whole fell from 829,000 to 801,000 and its share of all jobs from 9.8% to 8.1%. At the top of the same occupational class the opposite happened: business and administration specialists went from 357,000 to 707,000, a doubling, with accountants, management consultants, policy advisers, and HR specialists each growing around 100%. New vacancies for administrative personnel fell from 158,200 in 2022 to 114,500 in 2025, 27.6% fewer, against 8.6% fewer for all occupations. Those still in clerical work are old: 31.0% are 55 or over, against 23.3% of all workers. That makes a shrinking occupation a hiring problem for years to come.
