Room for Loss

The most productive document in Dutch health-care history is a plan that never made it. The Simons plan foundered in the early 1990s on political resistance, and without that foundering there would have been no Health Insurance Act (Zvw) in 2006.¹ Nº 00 called it the most expensive and most instructive phase of the design.² This paper is about the link that makes it possible, the first of two working forms in the chain: room for loss.

What room for loss is

Room for loss means that something is on the table which can be rejected, and that rejection settles something. That is the only environment in which the room for judgement around a system genuinely comes into being. As long as a proposal carries no commitment, no one has to make their trade-offs explicit; only when it can be rejected does the fight begin, and only in the fight does someone write down which trade-off was worth being defeated over. Nº 00 put it sharply: whoever stakes nothing loses nothing, and whoever loses nothing does not learn which trade-off was worth the fight.

The link must be distinguished from the fashion that seems, on the surface, to embrace it. The cult of failing fast celebrates the loss itself; this link is about what the loss leaves behind. An attempt that is defeated and leaves a decision point behind is an investment. An attempt that is defeated and leaves nothing behind is simply loss. The difference between the two is the legacy, and for that a word is needed which this series will go on to use: the inheritable failure.

The most expensive lesson that paid off

What exactly did the Health Insurance Act inherit from the foundered Simons plan? Three things, each obtainable only through loss. First, the knowledge of which transition was politically untenable: the route via predominantly income-related financing had been defeated in public, and that fight did not have to be fought again in 2004, because it had been settled. Second, the insight that the technique was not yet ripe: without a working risk-equalisation system there is no obligation to accept applicants, and that equalisation was developed further within the Sickness Funds Act (Ziekenfondswet) during the 1990s, in the shelter the failure had created.¹ Third, the map of the forces at play was settled for everyone after Simons. When the signatures were put to paper in 2006, the lost fights had already been fought.

The same mechanism works one level of scale up. The European Constitution was rejected by referendum in the Netherlands and France in 2005. The 2007 Treaty of Lisbon inherited the institutional substance and dropped the constitutional symbolism, precisely along the lines the rejection had drawn.³ At Union scale too, a rejected text leaves a decision point behind. A rejection is information of the highest quality; it is only not free.

Evaporating instead of being defeated

The empty case lies in education. The revision of the curriculum began in 2016 with the Onderwijs2032 advisory report, took a second form in 2019 as Curriculum.nu — the successive curriculum-revision programmes — and has run on since then under the banner of updating the core objectives.⁴ Three programme names, hundreds of teachers and experts involved, and only at the end of 2025, nine years after the first report, did something come up for a vote for the first time: an act revising the statutory foundations of the core objectives. The substance itself, the core objectives that the whole thing had been about all those years, is fixed outside that vote, by order in council. For nine years, then, nothing could be defeated, and the rounds show the pattern: every round met criticism, and every round responded by evaporating instead of being defeated. The programme was renamed, reshuffled, restarted.

Evaporating seems milder than being defeated and is in reality more expensive. A defeated proposal settles something: not this, and for this reason. An evaporated programme settles nothing; the next round begins with a new name and the same questions, and those involved in the previous round take their learned fights home with them. The tax dossier from Nº 00 follows the same law in another form: fifteen years of reports and no integrated bill that could be defeated. Where nothing can be rejected, nothing is settled.

The countable substitute

The substitute for this link is the pilot, with its family: the proeftuin (testbed), the living lab, the experimentation clause, the deal. The pilot is the attempt from which the loss has been surgically removed. It cannot be defeated: a successful pilot leads to exploring scale-up, a failed pilot is called a learning experience, and the end of the pilot is the end of the funding, rarely a decision. It is countable par excellence: launched, number of participants, lessons harvested, and those lessons then go into the knowledge base, which is where this substitute connects seamlessly to that of Nº 01.

None of this makes the experiment wrong as an instrument for learning. There are questions only a trial can answer. Substitution begins where the pilot takes the place of the proposal: where a system measures its courage in experiments launched and meanwhile puts nothing anywhere up for rejection. An administration that does nothing but pilot has abolished losing, and with it, unnoticed, settling.

Organising loss

The action perspective for this link demands administrative courage, but its form is a matter of craft.

The first move: let every system exploration end in a proposal that comes up for decision, with rejection as a real outcome. That sounds self-evident and no longer is; the usual route ends in a report, a cabinet response and a follow-up programme. The discipline is an end date with a vote.

The second move: make the legacy compulsory. Every programme that is defeated or discontinued closes with a short, public legacy document: what has been settled, what remained open, what has been learned. This is the instrument Nº 02 was pointing to when it spoke of inheriting. It costs an afternoon and it is the difference between a failure and an investment.

The third move: give every pilot a decision clause at the start. This trial ends on this date in a decision to introduce or to abandon, taken by this body. A pilot without a decision clause is not launched. Whoever introduces this rule will notice how many proposals for experiments suddenly stop coming, and that is the point: it filters the trial that wants to know something from the trial that wants to avoid something.

The test of this link: when did your organisation last put forward something that could be rejected, and what did that rejection leave behind? Whoever does not know the answer has not lost in a long time, and that is not a good sign.

A pilot ends with a report. An attempt ends with a decision. Only the second leaves anything behind.

Footnotes

1. System proposals based on the Dekker Committee (1987), foundered in the early 1990s (the Simons plan); further development of the risk-equalisation system within the Sickness Funds Act (Ziekenfondswet) during the 1990s; Health Insurance Act (Zorgverzekeringswet), in force 1 January 2006. ↩ ↩²

2. Jacob Huibers, Orphaned Design (Nº 00), Statecraft, 2026, statecraft.nl. ↩

3. Referendums on the Treaty establishing a Constitution for Europe, France and the Netherlands, May and June 2005; Treaty of Lisbon, signed 2007, in force 2009. ↩

4. Platform Onderwijs2032, final advisory report, 2016; Curriculum.nu, proposals, 2019; Wet herziening wettelijke grondslagen kerndoelen (Act revising the statutory foundations of the core objectives), adopted by the House of Representatives on 9 December 2025, before the Senate in mid-2026; the updated core objectives are fixed by order in council, intended from August 2026. ↩