The Headquarters Exodus: Why Shell, Unilever, DSM and Boskalis Left the Netherlands, Who Stays

The Netherlands has become the rare developed economy that simultaneously loses the head offices of its oldest multinationals and wins the legal homes of some of the world’s largest holding companies. Shell, Unilever and DSM, three names that defined Dutch corporate life for a century, moved their top holding company or their domicile out of the country between 2020 and 2023. Boskalis and Van Oord, the dredgers that built Dubai’s palm islands, shifted growth and part of their management to the Gulf. Yet in the same years Stellantis, Prosus, Exor and a long list of foreign groups chose a Dutch NV as their legal shell, and ASML, now the most valuable company in Europe, extracted a €2.5 billion package from The Hague in return for staying. This article explains the mechanics of the dividend tax that drove the departures, the politics that made Dutch policy feel unpredictable, and the 2025-26 reset under a new cabinet. It is part of the Netherlands Company Stories hub.

What does the “headquarters exodus” actually consist of?

The exodus is a cluster of five decisions between 2020 and 2024 in which Dutch-rooted companies moved their holding company, their tax residence, their listing or part of their management out of the Netherlands. Unilever unified in London in November 2020. Shell moved its headquarters and tax residence to London and dropped “Royal Dutch” from its name in January 2022. DSM merged with Firmenich in May 2023 into a group domiciled in Kaiseraugst, Switzerland. Boskalis was taken private by HAL Investments in 2022 and moved part of its head office to Abu Dhabi in 2024. Van Oord said in March 2024 that future expansion would happen from Dubai.

None of these companies closed a Dutch factory or sacked a Dutch workforce as part of the move. Shell still employs thousands at Pernis and Moerdijk, Unilever keeps its research campus in Wageningen, DSM-Firmenich opened a new Maastricht headquarters building on 30 May 2024 and Boskalis still runs its fleet from Papendrecht. What the Netherlands lost was the top of the corporate pyramid: the board meetings, the treasury, the tax base of the holding company and the symbolism of a Dutch blue-chip.

That symbolism matters more in the Netherlands than in most countries. Shell and Unilever were the two largest constituents of the AEX index for decades. Their departures were the most visible evidence for a thesis that PwC’s chief economist Barbara Baarsma repeated in September 2025: the Dutch business climate has worsened for three consecutive years and now scores below its 2013 level, when the country was still recovering from the financial crisis.

Why was the dividend withholding tax such a powerful push factor?

The Dutch dividend withholding tax is a 15 per cent levy deducted at source on dividends paid by a Dutch-resident company, and for Shell and Unilever it created an unequal treatment between two classes of shareholder in the same group. Both groups ran dual structures dating from the 1930s (Unilever) and 2005 (Shell). Dutch-registered shares bore the withholding tax; British-registered shares did not, because the United Kingdom levies no equivalent tax on outbound dividends.

For a pension fund in Texas or a retail investor in Singapore, this meant that a Shell A share paid out less cash than a Shell B share. Shell managed the gap with a “dividend access mechanism” that routed B-share dividends through a trust so they counted as UK-sourced. The arrangement worked but was fragile, attracted scrutiny from the Dutch tax authority and prevented Shell from merging its two share classes, which in turn made share buybacks and all-share acquisitions clumsier than they needed to be.

The Rutte III cabinet understood the problem and in 2017-18 proposed abolishing the dividend tax altogether, at an estimated cost of well over a billion euros a year. The proposal was deeply unpopular with voters, who saw it as a gift to foreign shareholders. When Unilever abandoned its plan to consolidate in Rotterdam on 5 October 2018, after British shareholders signalled they would vote it down, the government lost its best argument and dropped the abolition within days. From that moment, every Dutch multinational with a dual structure knew the tax would stay and that The Hague could not deliver on a promise made to business.

The lesson for the letterbox company debate was ironic: the same tax that the Netherlands used to defend itself against accusations of being a conduit for untaxed dividends was the tax that drove its own champions to London.

How did the 2020 exit tax proposal accelerate the departures?

In July 2020, as Unilever prepared its shareholder votes on unifying in London, GroenLinks MP Bart Snels filed a private member’s bill for a “conditional exit tax” that would have levied the 15 per cent dividend tax on a company’s retained earnings at the moment it moved out of the Netherlands. For Unilever the bill implied a potential charge measured in billions of euros. The company said it would not proceed with unification if the bill passed before completion.

The bill did not pass in time, Dutch shareholders approved the unification in September 2020 and roughly 99 per cent of British shareholders followed in October. The merger of Unilever NV into Unilever PLC completed in November 2020. Shell watched the episode closely. In November 2021 it announced it would move its headquarters and tax residence to London, abolish the A and B share classes and rename itself Shell plc. Shareholders backed the plan at a December 2021 meeting with overwhelming support, and the new name was registered at Companies House on 21 January 2022.

The exit tax episode mattered because it turned a tax disagreement into a question of trust. Company boards concluded that a future Dutch parliament could change the rules retroactively, and that the safest moment to leave was before any such rule took effect. That interpretation also shaped how the Shell move to London was explained to investors: as a simplification, not as a tax arbitrage, but with a clear subtext about predictability.

Why did DSM, Boskalis and Van Oord follow a different route?

DSM, Boskalis and Van Oord each left for reasons that had less to do with the dividend tax and more to do with merger logic, ownership and talent. DSM’s case was a merger: when it combined with Swiss fragrance and flavour house Firmenich on 9 May 2023, the new DSM-Firmenich took a Swiss domicile in Kaiseraugst while keeping a listing on Euronext Amsterdam and a major operational headquarters in Maastricht. The group reported revenue of €12.8 billion in 2024 and in February 2025 agreed to sell its share of the Feed Enzyme Alliance to Novonesis for €1.5 billion. Dutch employment and research were preserved; the legal seat was not.

Boskalis was a case of a long-term owner losing patience with public markets. HAL Investments, the investment vehicle of the Van der Vorm family, bid for the dredger in early 2022 at a valuation of about €4.3 billion, secured more than 95 per cent of the shares and delisted it from Euronext Amsterdam in October 2022. Once private, CEO Peter Berdowski moved part of the head office to Abu Dhabi in 2024, citing the availability of Asian engineering talent in the Gulf and what he described as a harsh Dutch policy climate for foreign labour. Boskalis reported revenue of €4.36 billion and net income of €781 million in 2024 with 11,683 employees, so this was a profitable company choosing where to grow, not a struggling one fleeing costs.

Van Oord, the family-owned rival, made the same choice in March 2024. Outgoing CEO Pieter van Oord told De Telegraaf that the company would remain a Dutch family business headquartered in the Netherlands, but that further expansion would take place from Dubai because of staff availability and tax. Van Oord had just reported a record 2023 turnover of €2.9 billion, up 42 per cent, with a net result of €127 million and nearly 6,000 employees, about half in the Netherlands. The two dredging giants thereby signalled that the constraint on Dutch growth was no longer capital or demand but people.

How did migration and expat policy become part of the business climate debate?

Migration policy entered the business climate debate in 2023-24 because the parties negotiating a new government all wanted to cut immigration, including the skilled migrants on which ASML, Boskalis and the Eindhoven chip cluster depend. The 2023 budget had already trimmed the “30 per cent ruling”, a tax break that exempts 30 per cent of a recruited expat’s salary from income tax, and the PVV and NSC campaigned for further limits on knowledge migrants and international students.

The reaction from business was unusually blunt for the Netherlands. ASML said it would consider expanding abroad if it could no longer recruit internationally. Eindhoven’s mayor Jeroen Dijsselbloem, a former finance minister, argued on television in March 2024 that every highly educated migrant ASML brings to Brainport creates jobs for three people with vocational training. Three University of Amsterdam professors wrote in Het Financieele Dagblad that unpredictability itself, even if measures were later reversed, was already damaging investment. DNB president Klaas Knot took a more nuanced line, acknowledging that skilled migrants add value while warning that the country was reaching limits in housing, grid capacity and nitrogen space.

By 2025 the political argument had shifted again. The partial reversal of the 30 per cent ruling cuts was enacted, but the Schoof cabinet collapsed in June 2025 when the PVV walked out, then again in August when NSC ministers left, leaving a VVD-BBB rump with 32 of 150 seats. Caretaker economic affairs minister Vincent Karremans told parliament in September 2025 that other European heads of state were personally telephoning Dutch technology companies to lure them abroad, and that the Netherlands needed a more active industrial policy than it was used to.

What is Project Beethoven and did it work?

Project Beethoven is a €2.5 billion public investment programme agreed in March 2024 to persuade ASML and the wider Eindhoven chip cluster to keep expanding in the Netherlands rather than abroad. The central government committed about €1.73 billion and Eindhoven with 20 surrounding municipalities, organised as Metropoolregio Eindhoven, added roughly €700 million. The money is earmarked for housing, road and rail infrastructure, grid capacity and technical education in the Brainport region, the bottlenecks ASML had named when it threatened to put new capacity elsewhere.

On the narrow question of whether ASML stayed, the answer is yes. ASML committed to expanding in Veldhoven and Eindhoven, and on 8 September 2026 it announced the start of construction of its “BIC North” campus in Eindhoven, with a first phase due by 2029. Its financial trajectory has made the stake higher: ASML reported net sales of €11.7 billion and net income of €4.96 billion in the third quarter of 2026, raised full-year guidance to €44-45 billion and told investors that even its planned capacity for 2027 and 2028 would not meet demand. The ASML EUV monopoly story explains why that demand is so inelastic.

Delivery has been slower than the headline. Omroep Brabant reported in January 2025 that smaller municipalities lacked the civil servants to execute the housing and infrastructure plans, with an official in Best expecting delays of at least a year. Karremans said in September 2025 that he received monthly updates on Beethoven, personally asked other grid users to wait so ASML could connect, and still could not guarantee permits. GroenLinks-PvdA MP Joris Thijssen summarised the risk: if grid capacity and nitrogen permits are not resolved, ASML cannot expand regardless of the money.

Which companies still choose a Dutch NV, and why?

Despite the departures, the Dutch NV remains one of the most used holding structures in Europe, because Dutch company law offers flexibility that French, Italian and German law do not. Stellantis, formed in 2021 from Fiat Chrysler and PSA, is a Dutch NV with its registered seat in Hoofddorp, even though its CEO Antonio Filosa, appointed in June 2025, runs the group largely from Auburn Hills, Michigan, and its 2024 revenue of €156.9 billion was earned mostly in North America and continental Europe. Exor, the Agnelli family holding that owns 15.5 per cent of Stellantis, is also Dutch-incorporated.

Prosus, the Amsterdam-listed technology investor, is the Netherlands’ largest case of inbound domicile. Its South African parent Naspers chose Amsterdam for a September 2019 listing that was valued at around €120 billion on its first day, and Prosus has since used its Dutch base for the €4.1 billion takeover of Just Eat Takeaway completed in late 2025. The Prosus story covers the mechanics of that structure.

The attraction is threefold. First, Dutch law permits loyalty voting shares, which let founding families such as the Agnellis or Peugeots keep control with a minority economic stake. Second, the Netherlands has more than a hundred bilateral tax treaties and a participation exemption that makes it efficient to collect dividends from subsidiaries worldwide. Third, Amsterdam is a liquid euro-denominated market with an English-language corporate culture. None of these advantages depends on the dividend tax, because a holding company that reinvests rather than distributes to Dutch-taxable shareholders feels the tax only lightly.

What role does the share buyback exemption play in the 2025-26 debate?

The buyback exemption is the second front in the dividend tax war, and it illustrates how Dutch policy swings. Under Dutch law a listed company that repurchases its own shares does not have to withhold dividend tax on the amount paid out, provided certain conditions are met. The exemption was meant for occasional buybacks; it has become structural. Research group SOMO calculated in October 2025 that Dutch listed companies spent €102 billion on buybacks between 2021 and 2024, more than French or German companies, and that abolishing the exemption could raise at least €2.2 billion a year. The Central Planning Bureau had earlier estimated the cost at about €1 billion.

In 2023 the Tweede Kamer voted to scrap the exemption from 2025 to fund a higher minimum wage and child benefits. In 2024 the new Schoof coalition reversed that decision before it took effect, after lobbying from VNO-NCW and warnings that Shell-style departures would follow. SOMO’s analysis of the 2025 election programmes found that VVD, CDA, D66 and most other parties wanted to keep the exemption, while GroenLinks-PvdA, ChristenUnie and Volt wanted to end it.

The reversal was good news for boards but reinforced the perception of a policy that is decided and undone every eighteen months. France introduced an 8 per cent buyback levy in 2025 on companies with revenue over €1 billion and the United States has charged 1 per cent since 2023, so the Netherlands is now more generous on buybacks than its peers, even as the 15 per cent dividend tax remains in place. Separately, the finance ministry has been investigating tax-free capital repayments after NRC reported in April 2024 that chemicals group OCI had returned about €5 billion to shareholders through capital reductions and avoided roughly €750 million in dividend tax, a route also used by AkzoNobel, TomTom and Ahold.

How did the 2025 election and the Jetten cabinet change the outlook?

The 29 October 2025 election produced a fragmented result in which D66 and the PVV tied on 26 seats each, the VVD took 22, GroenLinks-PvdA 20 and the CDA 18, and the eventual outcome was a centre to centre-right minority cabinet of D66, VVD and CDA under Rob Jetten, sworn in on 23 February 2026. It is the first minority cabinet in the Netherlands without a formal confidence-and-supply arrangement since 1918, which makes long-term tax promises harder to bank, but its composition is the most business-friendly in a decade. Eelco Heinen (VVD) continues as finance minister, Heleen Herbert (CDA) holds Economic Affairs and Climate Policy, Sjoerd Sjoerdsma (D66) is foreign trade minister and Eelco Eerenberg (D66) is state secretary for tax.

For business the relevant signals are the absence of the PVV and NSC, which drove the 2024 migration restrictions, and the presence of D66, the party that has most consistently argued for skilled migration and research spending. Sjoerdsma travelled to Beijing in July 2026 to defuse the Nexperia and ASML disputes with China, a topic covered in the de-risking story. The cabinet has also inherited the Beethoven commitments and the pending question of the buyback exemption.

What the cabinet has not done, as of October 2026, is touch the 15 per cent dividend withholding tax. Abolition remains politically toxic after 2018, and the fiscal room is limited by defence spending and the pension transition. The practical consequence is that the Netherlands will continue to be a good domicile for holding companies and a difficult one for dual-listed groups with large British or American retail shareholder bases, exactly the pattern of the last six years.

What can founders and investors learn from the Dutch headquarters exodus?

The main lesson is that domicile decisions are driven by the interaction of tax design and political credibility, not by headline tax rates. The Dutch corporate income tax rate of 25.8 per cent is unremarkable in Europe, and none of the companies that left cited it. What they cited was a specific withholding tax that penalised one class of their shareholders, and a legislature that twice changed course on it within two years.

For founders building a European group, the Dutch experience argues for choosing a legal seat early, before a dual structure or a large cross-border shareholder base makes a later move expensive and politically charged. Stellantis, Prosus and Exor made that choice at formation and have had no reason to revisit it. Shell and Unilever inherited structures from the 1930s and 2000s and paid the price in years of management time.

For CFOs, the buyback exemption and capital repayment rules show that cash return mechanics can be worth hundreds of millions of euros a year, but that the legal basis can shift with each coalition. Building a distribution policy that works under both the current rules and a plausible reform scenario is cheaper than being forced into a defensive redomiciliation.

For investors, the Dutch case is a reminder that governments can respond. Project Beethoven, whatever its execution problems, demonstrates that a state will spend €2.5 billion to keep a company whose quarterly profit now approaches €5 billion. The companies most likely to receive that treatment are those with scarce technology and visible employment, as the Dutch chip cluster illustrates; consumer goods and energy groups with mobile headquarters were allowed to go. Readers comparing jurisdictions will find a parallel debate about headquarters flight in the UK Company Stories hub.

Frequently Asked Questions