Loopholes and Thresholds

A municipality wanted to know which children came to school without breakfast. The reasoning was straightforward: an empty plate points to poverty, poverty correlates with low income, so a breakfast provision at the schools in the poorest neighbourhoods would reach the children who needed it. The provision came, and with the provision came a measurement, because that is how a pilot works. And the measurement pointed at something that did not fit the reasoning. A considerable share of the children without breakfast turned out to come not from the poorest families but from households with two working parents. It was not the money that was missing but the morning: two parents out of the door by a quarter past seven, a child fending for itself, a plate that stays empty for reasons that have nothing to do with income.

That is a rare moment, and it is worth pausing on before it gets organised away. Usually a measurement fits neatly onto the assumption that ordered it, and nobody notices that the assumption narrows reality. Here the measurement ran ahead of the assumption and contradicted it. The empty plate — the countable quantity — turned out not to correspond to the need it was supposed to measure. The proxy caught itself out.

This piece is about what happens when a proxy does not catch itself out. It is about the needs that the administrative system can only see when they manage to attach themselves to an existing measurement, and about the people who drop out of view because their need cannot. And it is, in the second half, about a symmetry that becomes visible as soon as you take countability seriously: the same fragmented architecture that lets wealth through at the top of the distribution holds income back at the bottom. Loopholes above, thresholds below, and the same hole in the design making the difference.

Which need gets a name

Not every need gets a word that sticks. Some concepts land, attach themselves to a measurement, unlock a provision and become part of the administrative vocabulary. Others float, however sharp they are, and disappear again. The difference between the two says less about the needs than about the carrier that must be able to hold them.

Take the precariat. It is one of the sharpest concepts offered in the past fifteen years for what many people at the bottom of the labour market actually experience, and it has never taken institutional root in the Netherlands. That is striking, because the word exists, it is well founded, and it names something real. The explanation does not lie in ignorance. The precariat describes existential insecurity: not income at a measuring moment, but the absence of security across time — no permanent contract, no predictable path, no anchoring of existence. That is, by construction, a quantity the system does not measure. The system counts income at a moment, labour participation at a moment, education level at a moment. Insecurity across time is not an entry in that ledger. A concept that lands nowhere countable keeps floating, however true it is.

Compare it with concepts that did land. The unemployed person landed, because a benefit hung from the word and a counter kept track of him. The occupationally disabled person landed for the same reason. Even the social assistance recipient is an administrative category before being a sociological one. Those words gained a foothold because a carrier lay beneath them that made them countable and attached a money flow to them. The precariat has no administrative twin, and in a system that only sees what it counts, that is a death sentence for a word.

There is a second way a concept can fail, and it is almost the mirror image of the first. The rugzakje (“little backpack”) in education — a personal budget attached to a child with additional needs — landed not too badly but too well. It became countable, it unlocked a budget, the counter started running, and when a large share of pupils turned out to justify some form of extra support, the exception had become the norm and the logic of exception unaffordable. It did not disappear because the need disappeared. It disappeared because the measurement registered the need too well. Passend Onderwijs (“Appropriate Education”, the 2014 reform) was the recoding that made the visible invisible again: the entitlement dissolved into a duty of care held by the school, and the counter disappeared with it. The precariat fails at the bottom end of countability, the backpack at the top end. The one never becomes a budget line; the other becomes such a large line that the system closes the line. Both end in the same place: a real group the system no longer sees.

Between those two extremes lies a third mechanism, and it is the most treacherous because it looks like success. Period poverty landed, and transport poverty landed partially, in both cases because they could attach themselves to a countable sub-element. A pack of sanitary towels costs an amount, that amount can be counted, so the need got a name and a distribution point. A bus ticket costs an amount, so mobility poverty became the absence of that amount. But note what lands and what does not. What lands is the sub-element, not the need beneath it. A girl who cannot afford sanitary towels presumably cannot afford more things, but only the one countable chunk gets a provision. Mobility poverty is not the absence of a ticket; it is the absence of reach, and reach depends on where the stop is, whether the line still runs, whether the amenity has been withdrawn from the neighbourhood. By narrowing the need to the countable sub-element, the actual need disappears from view once more, and a spatial question is reduced to an individual ticket question.

That is the proxy that does not catch itself out. With the school breakfast, the measurement contradicted the assumption. Here it does not. The counter falls, the provision reports progress, and the system registers success, while the need beneath remains untouched — because it was never the plate or the pack or the ticket. The measurable shadow of the need is mistaken for the need itself.

The pattern behind these cases is more precise than the observation that concepts sometimes fail to land. A concept that wants to make a need visible survives only if it can attach itself to an existing measurement, without threatening the premise of the system, and without placing the cause outside the individual. The precariat does not attach and floats. The backpack attaches too well and is coded away. The working poor — the concept for those who work and are poor anyway — threatens the premise that work leads out of poverty, and is therefore pushed away into the individual case instead of adopted as a category. The carrier admits one type of need: countable, affordable, premise-confirming. Everything that does not meet those conditions the system does not see, and neither, therefore, the people behind it.

The silence is not an accident

Whoever talks a lot with people at the bottom hears the crumbling before seeing it measured. That is not an anecdotal shortcoming of the figures; it is the core of the problem. The figures are not silent because there is nothing; they are silent because they do not measure it. The silence around lived decline is not the absence of the decline; it is the absence of a register in which the decline can report itself. A carrier that could hold the lived register would not be silent. It would sound the alarm.

That silence has a price at the receiving end that rarely comes into view. Because the system can only see the need in countable fragments, it also compensates it in fragments. For the empty plate a breakfast provision, for the pack of sanitary towels a distribution point, for the ticket a reimbursement, and alongside them special municipal assistance, the child package, the youth sports fund, the appliance scheme, the computer scheme, the individual income supplement. Each of those schemes is, in itself, a reasonable answer to a real sub-element. The sum is something else. Whoever lays the schemes side by side sees not a collection of solutions but an imprint: a negative image of a need that is nowhere recognised as a whole. Every column has found its own countable chunk of the poverty and built a little pot around it. Nobody carries the whole, because the whole is countable in no column.

How that fragmentation between budgets produces perverse outcomes that land selectively where counter-force is missing, is worked out elsewhere in this corpus. What counts here is the flip side on the citizen’s end. The self-reliance the whole system presupposes is precisely the competence needed to know, apply for, prove and account for dozens of application streams, each with its own counter, its own income test and its own risk of clawback. That competence is most often missing in those who need the schemes most. The fragmentation thereby produces non-take-up, and the non-take-up is invisible once again, because a pot that is not applied for appears in no counter as need. The carrier registers the expenditure, not the missed entitlement. Whoever is not reached does not exist for the system.

Up to here the gaze has been on the bottom. But the architecture that fragments the bottom is the same one that fragments the top, and there it becomes visible in its opposite.

At the top of the wealth distribution, what fragmentation does there is documented. The highest incomes carry a lower average tax burden than the middle groups, and the business succession deferral scheme yields effective rates on wealth transfers of a few per cent for those able to use it.¹ The system is progressive on paper and not in practice. The mechanics are familiar: whoever has the possibility of placing their wealth where the lowest burden applies does so, and the fragmentation of the system across tax boxes, schemes and charging moments supplies exactly the openings through which that can be done. Here we call those loopholes. A loophole is an opening in the fabric you slip through if you know how.

At the bottom of the income distribution, the same fragmentation produces the reverse. Whoever moves from benefits into work, or from part-time into more hours, sees part of the income rise leak away as allowances and schemes fall off. The ministry measures this as separate thresholds: the unemployment trap, the re-entry trap, the part-time trap, each measured in terms of effective marginal pressure — and for some household types around the minimum wage that pressure runs high through the simultaneous phase-out of the child budget and the housing benefit.² Here we call those thresholds. A threshold is an opening in that same fabric you trip over if you do not see it coming.

Now the thesis this piece arrives at. The loophole and the threshold are not two different systems, one for the rich and one for the poor. It is the same hole in the same architecture. The difference is not in the opening; it is in the sight of whoever encounters it. A loophole is an opening seen by someone with the means to see it: the adviser who calculates the outcome in advance, the private company in which the wealth waits, the timing that can be planned. A threshold is the same kind of opening seen by someone without those means: a household that cannot survey its marginal pressure in advance, that gets a different outcome per scheme, per moment and per possible error, and that therefore trips where the other slips through.

That explains the two corners that close the thesis. It is not that the top encounters only loopholes and the bottom only thresholds. The top could trip over the same thresholds, but does not, because wealth buys the sight that turns the threshold into a loophole. The bottom could use the same loopholes, but cannot, because it lacks the sight to find them. Minimal pressure where maximal wealth sits, maximal pressure where minimal income sits — and the hinge variable is not the opening but the sight, which is unequally distributed along precisely the axis of wealth.

Here an objection presents itself, and it makes the thesis stronger rather than weaker. Research commissioned by the Ministry of Finance has shown that people on low incomes often overestimate their marginal pressure, on average by tens of percentage points, and that the actual pressure for many workers is lower than feared.³ One might conclude that the threshold is partly imagination. But look at what the system does with that finding. It establishes that the lived pressure is higher than the measured pressure, and it decides that the lived pressure is the error to be corrected — through information campaigns, a letter, the message that working more pays better than people think. The perception is declared the problem, not the architecture that makes the perception entirely reasonable. For why would someone with twelve income-dependent schemes, each of which can fall away at its own moment, and with the knowledge that one error produces a clawback, not assume the worst? The overestimate is not a misunderstanding. It is a correct assessment of the risk of a system you cannot survey.

With that, the symmetry shifts from amounts to predictability, and becomes more robust than any percentage. The top does not merely have low pressure; the top has certainty: the outcome is calculable in advance, so the loophole is usable. The bottom does not merely have high pressure; the bottom has uncertainty: the outcome is unknowable for the household itself, so the threshold is unavoidable. Predictability is precisely what makes adaptation possible. Whoever knows the outcome can arrange themselves around it. Whoever does not behaves like the prisoner of a trap that is partly real and partly unknowable — which in practice comes down to the same thing.

And so the piece closes where it opened. The information campaign against the overestimated marginal pressure is the same mechanism as the school breakfast, only on the remedy side. The need is unpredictability, the countable sub-element is the measured average pressure, and the campaign fights the countable element while the need — the inability to survey — remains untouched. The system cannot measure the uncertainty it itself produces, so it redefines it as a perception problem of the citizen, and sends a letter. Proxy substitution at the level of the solution.

The question that remains

The obvious conclusion is that the thresholds must be demolished and the loopholes closed, and that is not wrong, but it is the wrong place to end. A system that replaces the twelve little pots with one bare amount closes the hole by abolishing one register: the normative register that sits in the fragments, the recognition that a child ought to have breakfast, to play sport, to cycle. That register disappears with a line through the pots. Closing the loopholes at the top and levelling the thresholds at the bottom is, moreover, a distribution question, and that question needs an owner with a mandate that Dutch government on this terrain precisely lacks.

What does lie within reach is the gaze itself. The carrier sees the same opening for the loophole and the threshold, and does not register who can or cannot get through. It measures the expenditure but not the missed entitlement, the average pressure but not the unpredictability, the empty plate but not the reason it is empty. The first step is not rebuilding the system; it is arranging the carrier so that it holds its own openings against reality, as the breakfast pilot accidentally did. A measurement that holds the lived register without reducing it to what is already counted is no remedy for the poverty. But it is the condition under which the system could for the first time see the poverty as a whole, instead of as the sum of its compensated fragments.

The question, then, is not whether the ordinary man is worse off. The question is why an administrative system with rising figures structurally cannot see that it is crumbling on the ground, and what kind of carrier is needed to make the lived register countable again without betraying it. As long as that carrier is missing, the silence remains. Not because there is nothing to hear, but because there is nothing that listens.

“Loopholes and Thresholds” is a Statecraft essay in the series of standalone publications, alongside the connected series that carry the corpus. It builds on the commentary “The Sum Has a Name” (May 2026) and on paper IV of Series II, “The Pressure on the Weakest”, and asks the question left unanswered in both: why the system cannot see the outcomes it produces as need.

Jacob Huibers is an interim manager with more than twenty years of experience in the Dutch public sector. He has worked as cluster manager, cluster director and quartermaster at municipalities ranging from fifty thousand to over two hundred thousand inhabitants, and at inter-municipal collaborative bodies across the social and physical domains. Statecraft is his platform for strategic reflection on public-sector execution, pillar IV of House of Viridian.

Responses and counter-arguments via Statecraft.nl.

CPB Netherlands Bureau for Economic Policy Analysis, De hoogste bomen vangen minder wind: belastingdruk op inkomens en vermogens (The tallest trees catch less wind: tax burden on incomes and wealth, May 2026). The effective rates on wealth transfer via the business succession deferral scheme are estimated there at a few per cent for those able to use it, against an average burden of thirty to thirty-five per cent for the middle groups. See also the discussion in the Statecraft commentary The Sum Has a Name (May 2026). ↩
Dutch Ministry of Finance, Tabellen marginale druk (Tables of effective marginal pressure) accompanying the 2025 Tax Plan (September 2025), and the explanation of the unemployment trap, re-entry trap and part-time trap in the budget documents. Effective marginal pressure varies strongly by household type and income trajectory; for single parents around the minimum wage it runs up through the simultaneous phase-out of the child budget and the housing benefit. ↩
Erasmus University and the Dutch Ministry of Finance, research into misperceptions of marginal pressure (reported October 2025). Households with a gross monthly income below two thousand euros overestimate their marginal pressure by on average roughly twenty-eight percentage points. The policy response focuses on information, including the working-hours calculator of Nibud, the national household finance institute. ↩

Which need gets a name

The silence is not an accident

Loopholes and thresholds

The question that remains

Footnotes