The Sum Has a Name

The CPB confirms the pattern, not the architecture that produces it

A response to the report by the CPB Netherlands Bureau for Economic Policy Analysis, "De hoogste bomen vangen minder wind: belastingdruk op inkomens en vermogens" ("The Tallest Trees Catch the Least Wind: the tax burden on incomes and wealth"), May 2026. The title inverts a familiar Dutch proverb that holds that tall trees catch the most wind, meaning that those at the top bear the heaviest scrutiny. The CPB's finding is the opposite.

The CPB, the Dutch government's economic policy bureau, published a report yesterday that is in every respect a statecraft document, except in name. It shows that the top 0.01 per cent of incomes carry an average tax burden of 28 per cent, while middle and upper-middle incomes pay between 30 and 35 per cent.¹ The system is progressive on paper and regressive in practice. It does not slow the growth of income and wealth disparities, and at certain points actively contributes to them. The wealth position of children is increasingly correlated with that of their parents, particularly at the top: children of parents in the top 0.01 per cent end up, on average, in the 96th percentile of their own generation. Equality of opportunity is not declining despite the tax system, but partly through the way it has been built.

So far, the facts. The question for readers of Statecraft is a different one: how does this report fit into the broader pattern we have diagnosed elsewhere, and why does the CPB itself remain inside the very register that keeps that pattern in place?

The pattern the CPB describes

In an earlier Statecraft piece, "The Invisible Policy", the mechanism was demonstrated across six sectors of the Dutch economy: housing, agriculture, primary healthcare, childcare, holiday-park real estate, and infrastructure. In each case the same sequence: an asymmetric impact on costs, an outflow of private and family-owned operators, an inflow of institutional or foreign capital, and irreversible entrenchment by the time the pattern reaches the talk shows. The CPB report adds a seventh case to that catalogue, and it is the most fundamental one. Not the distribution of wealth within a sector, but the distribution of the tax burden itself.

The mechanics are identical. Wage earners cannot defer; sole traders cannot defer; majority shareholders of closely held companies, known in Dutch as directeuren-grootaandeelhouders or DGAs, can. Liquid wealth in Box 3, the part of the Dutch personal income tax that covers savings and investments, is taxed immediately. Wealth held inside a private limited company falls under Box 2, where retained earnings are not yet taxed and the eventual claim can be deferred almost indefinitely through the doorschuifregeling, a rollover provision originally introduced to protect family businesses from forced sale at the moment of succession. The effective rate paid on transfers benefiting from this combination of provisions averages roughly 2 per cent on gifts and 5 per cent on inheritances among those who use them.² That is not an unforeseen side effect; it is what the design produces. Anyone able to place their wealth where the burden is lowest does so. Loans drawn from one's own private company doubled to 65 billion euros between 2010 and 2020. Substantial-interest wealth held in private limited companies grew by a factor of two and a half between 2006 and 2024, while Box 3 wealth contracted in real terms.

The CPB writes it in the careful prose available to a planning bureau: "the largest categories of wealth carry the lowest tax burden and have grown the most".³ In Statecraft terms: arbitrage at the design stage is invisible, because the design itself emerges in fragments and no one carries the mandate to assess the sum. Box 3 sits at the Ministry of Finance. Corporation tax sits at Finance. Inheritance and gift tax sit at Finance. Owner-occupied housing provisions sit between Finance and Interior Affairs. The pension framework sits at Social Affairs. On paper this is all the same government, yet the modelling assumes symmetric actors who in reality are profoundly asymmetric.

What the report sees and what it cannot see

This is where the CPB runs up against the limits of its own institutional position. As a piece of work in the rational-planning register, the report is excellent. It quantifies, it traces the mechanics, it sketches policy options. What it does not do, and cannot do from where it sits, is name the architecture that produces the sum. The report speaks of "trade-offs" and "distortions", of "effectiveness" and "efficiency". This is the blue language of the building-block studies and inter-departmental policy reviews that populate the Dutch policy ecosystem. In the Change Colours framework of De Caluwé and Vermaak, widely used in Dutch public-sector reform, blue is the register most often deployed in public transformations and the one that performs least well precisely on questions of power.

The yellow corner, the corner of power and partisan interest, is touched once in the report. The CPB notes that economic concentration may translate into the use of political influence to serve sectional rather than general interests.⁴ That is not a footnote; it is the core. A tax system in which the levers of deferral and arbitrage are concentrated among roughly fourteen hundred people at the top, and in which the fiscal facilities for owner-occupied housing and pensions accommodate primarily middle and upper-middle incomes, is not the accidental outcome of technical decisions. It is what remains after decades of separate instrument choices have summed in a field where those best able to adapt to each new rule have done so. The CPB describes this with full propriety, and is then unable to frame it as a failure of statecraft, because in the Netherlands statecraft as an administrative capacity has no owner.

The unintended foundation

For readers familiar with the concept of borging, the term used in the Statecraft series for the test of what remains standing once no one is tending it any longer, this dossier offers an instructive inversion. Borging usually concerns the question of how to anchor public value so that a reform survives the departure of its programme manager or a shift in political wind. Good administration anchors public value in people, processes, financing and decision structures, so that the work does not unravel the moment attention moves elsewhere. In this case the tax system performs a different kind of anchoring, and it performs it with remarkable efficacy. It anchors the position of the wealthiest, generation after generation, through a mechanism that no founding document spells out as a goal. The rollover provision was not designed to allow wealth to pass untaxed across generations. But that is what it produces after three generations have used it. Between 2006 and 2022 the median wealth of the median household rose from 47,000 to 135,000 euros; the median wealth of the top 0.01 per cent rose from 52 to 62 million. Anyone applying the test of what remains standing once no one is steering finds an exceptionally well-anchored result, only not for the value that the legislator professed in the plenary debate.

The Strategic Triangle

In the terms of Mark Moore's Strategic Triangle, this dossier is an example of a triangle whose three corners each lie in their own plane. The public value, taxation according to ability to pay and reasonable equality of opportunity, is endorsed across the political spectrum in language each party finds congenial. The operational capacity, in the form of the Dutch Tax Administration, has been overstretched for years by precisely the complexity that each of those parties has added through amendments and special provisions. The political legitimacy for structural adjustment is captured by the organised interests of the groups that benefit most from the current asymmetry, which are, as a rule, better organised than the groups that pay for it. The three corners exist, but they do not close. Anyone wishing to bring the system back into balance cannot strengthen one corner. He must bring the triangle back into a single common plane, and that requires something rarely available in The Hague: an institution with the mandate to assess the sum, a horizon of fifteen to twenty years, and the willingness to name the yellow corner explicitly.

What this means for administrative practice

The CPB report will be discussed in the customary way over the coming weeks: in standing committees, in opinion columns and editorials, in an electoral programme that adopts a few of the recommendations and quietly leaves the difficult ones aside. For those working in the social or physical domain of a Dutch municipality, the relevance of this report is indirect but real. The recurring municipal funding shortfalls known as ravijnjaren, literally "ravine years", a term that emerged after the 2015 decentralisation of youth care, long-term care and labour-market activation to local government, when central transfers structurally fell short of devolved responsibilities; the deficits in youth care; the WMO thresholds, the long-term Social Support Act under which municipalities deliver community care, that are tightened year on year; the executive capacity that erodes in plain sight: all of this is enforced within a fiscal framework whose structural skew at the top is left untouched. It is no accident that central government imposes ravine years on municipalities while leaving the fiscal headroom at the top unused. It is a design choice, even though no one took it as such. The sum chooses, and the municipality pays for the execution.

In Chapter 9 of De Richting van de Beweging (The Direction of the Movement), the manuscript in preparation in which the four core models of the Statecraft corpus are worked out, this is developed further as the problem of transferwaarde, the degree to which a reform stands on its own once the reformer has gone. The present dossier illustrates the inverse. A system can also, intentionally or not, anchor the absence of reform, and it can do so through the same mechanisms that earn praise in other fields. Continuity, predictability, compatibility with existing structures. What functions as a virtue at one level functions as a bolt at another.

Statecraft, in this light, is the discipline of seeing both levels at once.

---

Footnotes

1. CPB, De hoogste bomen vangen minder wind: belastingdruk op inkomens en vermogens (May 2026), p. 11. The figures derive from Van Essen, Vanheukelom, Schulenberg and Lejour (2024), Inkomens en belastingen aan de top in Nederland ("Incomes and taxation at the top in the Netherlands"), CPB. ↩

2. CPB (2026), p. 19; the average effective tax rate on declared inheritances and gifts is taken from Möhlmann (2022), Evaluatie fiscale regelingen gericht op bedrijfsoverdracht ("Evaluation of tax provisions for business succession"), CPB. On the correlation between parental and filial wealth, see Schulenberg, Van Essen, Hamelink and Lejour (2024), Een steuntje in de rug: vermogensmobiliteit van ouder op kind ("A helping hand: intergenerational wealth mobility"), CPB. ↩

3. CPB (2026), p. 17. Underlying figures on the development of substantial-interest wealth between 2006 and 2024 derive from Statistics Netherlands (CBS) data cited at pp. 17-18. ↩

4. CPB (2026), p. 21, box 3, citing Banerjee and Duflo (2003), "Inequality and Growth: What Can the Data Say?", Journal of Economic Growth 8: 267-299. ↩