The Optimisation Asymmetry — Statecraft

§ 01 · A house with the meter at zero

In a Drenthe village stands a row of houses with a glossy completion certificate. Net-zero-on-the-meter (nul-op-de-meter), delivered by a housing association together with a major contractor, financed through an Energieprestatievergoeding (Energy Performance Allowance), completed in the years when Stroomversnelling (the Dutch energy-renovation programme) was being held up as a national exemplar.¹ In one of those houses lives a woman who has told her story to the regional press. It sounded marvellous, she says in the newspaper article, *a renovated and improved home for the same money. But if I had known what was waiting for me, I would not have done it. It is a nightmare.*²

The meter reads zero. The system works. The energy balance is correct, the subsidy conditions have been met, the inspection report is green.

A second blanket lies on the sofa. The heat pump hums; the indoor temperature hovers around seventeen and a half degrees. The MVHR unit can be heard in the bedroom; the resident sleeps worse than she did before the renovation. Mould creeps along the edges of the bathroom ceiling. Condensation lines the kitchen wall. Outside, the winter sun runs across the façade panels. Inside is a house which to the meter is a house, and which to its inhabitant is no longer one.

The meter reads zero. Someone lives in that house.

The pattern this paper names sits in the relationship between those two sentences. The meter has been carefully designed. The dwelling has been carefully tested. There is no carelessness in play, no bad faith, no incompetence. The optimisation has succeeded. What has succeeded is precisely one variable, and the other variables that previously belonged to the concept house have not so much failed as been removed from the system. They are no longer part of the definition. They have become waste of a process that was aimed at something else.

§ 02 · What the pattern is, and what it is not

The optimisation asymmetry is not a conspiracy. Not stupidity. Not a lie. It is a specific form of structural distortion that arises when a complex, multidimensional process — a house, a neighbourhood shop, a senior civil-service post, a therapeutic relationship, a consultation room — is governed by a single measurable variable, and when that one variable is tied to reward or sanction. The other dimensions of the process, dimensions that were constitutive of the primary work but are not, or only poorly, measurable, are then expelled from the system as a by-product. Not because anyone wished to remove them. Because in the operational logic of the system they no longer have a role.

Three elements together produce the pattern. First, a choice is made to render one dimension operational, that is, convertible into figures, comparable across cases, verifiable by a third party. Then a target value is attached to that dimension, a norm by which the success of the intervention is read off. Finally, the target value is bound to reward or sanction: in financing, in career prospects, in public reputation, in the granting of permits. From that point on, the optimisation spiral runs predictably.

The pattern is wider than the public-sector domain alone. It appears in any organisation in which accountability is rendered through figures and in which those figures carry consequences. In the public sector the pattern is more sharply visible because the goals are multiple (public value, operational capacity, political legitimacy, in the terms of Mark Moore’s Strategic Triangle³) and because accountability for one of those goals almost always comes at the cost of the other two. An intervention that optimises for operational capacity (cost price, throughput, processing time, mobility) almost certainly cuts something away from public value or political legitimacy. The three corners are seldom in balance, Moore wrote already in 1995. The optimisation asymmetry is what happens when balance is no longer sought, and one corner has taken the whole table.

Three clarifications belong here. The pattern is not the same as measuring incorrectly. The variable is often very carefully measured, and the measurement is technically correct. The problem does not lie in the measurement but in the choice to make that one variable into the goal in a system whose value is multiple. The pattern is not the same as measuring too little. More measurement does not solve the asymmetry; it relocates it. When new indicators are added to a one-sided KPI dashboard, the new indicators are themselves optimised in turn, and what falls between the meshes of the expanded set becomes waste once again. The pattern is not the same as choosing the wrong variable. Often the chosen variable is in itself meaningful. The energy balance of a dwelling is a legitimate indicator. The pick rate in a distribution centre is an operational reality. The unit cost of a care trajectory is a budgetary necessity. The pattern arises not from a wrong choice; it arises from a chosen variable growing, without counterweight, into the centre of the system.

The regularity behind it is well known and bears three names, all formulated in the second half of the twentieth century by authors who had not read each other.

§ 03 · Goodhart, Campbell, Strathern

Charles Goodhart, a monetary economist at the Bank of England, observed in 1975 that as soon as the British monetary authority elevated a statistical regularity in money-supply growth to the status of a policy target, that regularity lost its predictive power.⁴ His formulation reads: any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes. Market actors adjust their behaviour as soon as a particular aggregate acquires governmental significance; the old correlation between aggregate and underlying economic reality breaks down. Goodhart’s observation was technical and local. It applied to monetary policy. But its reasoning was more general.

Donald Campbell, a social psychologist at Northwestern University among others, formulated in 1979 a regularity that ran in parallel but came from the other side. *The more any quantitative social indicator is used for social decision-making, the more subject it will be to corruption pressures and the more apt it will be to distort and corrupt the social processes it is intended to monitor.*⁵ Campbell’s empirical examples came from American education and crime statistics: as soon as school tests formed the basis for funding, schools learned to teach to the test; as soon as arrest figures became the leading chapter in the police annual report, arrest behaviour shifted in the direction of what could be counted. Campbell’s contribution was to show that the distortion was not a marginal phenomenon but a property of the system. Anyone who ties a social indicator to decisions sooner or later acquires the distortion as part of the bargain.

Marilyn Strathern, an anthropologist at Cambridge, condensed both observations in 1997 into a single line that has since become the signature of the pattern. In an article on the British university Research Assessment Exercise she wrote: when a measure becomes a target, it ceases to be a good measure.⁶ Strathern carried the observation into the audit society that had developed around 1980 in Western Europe and North America. Universities, hospitals, municipalities, child-protection providers, housing associations: in all those domains, ranking, scoring and target systems were introduced over the course of two decades, frequently with the explicit ambition of fostering quality. Strathern’s claim was that the mechanism Goodhart had seen in monetary policy and Campbell had seen in education was equally at work across all those domains. As soon as a measurement becomes a goal, it loses its diagnostic value for what it had first set out to measure.

Three authors, three domains, one pattern. The regularity has since been confirmed in countless contexts, most authoritatively by Wendy Espeland and Mitchell Stevens in their work on commensuration and quantification, by James C. Scott in Seeing Like a State on the legibility that the modern state demands of its population, by Michael Power in The Audit Society on the autonomous logic that auditing develops, and by Jerry Muller in The Tyranny of Metrics on what happens when education, healthcare, policing and the military fall into metric fixation.⁷

The shared core in all those works is the same. What is measurable becomes visible; what is visible becomes governing; what is governing displaces what is not measurable; and what is not measurable undergoes the displacement without any deed of banishment ever being drawn up. The becoming-waste is silent. It does not happen by decision. It happens through the presence of a dominant variable that slowly drives all other variables out of the operational description of the work.

In the Dutch public sector, the pattern is at work in many places. Five domains are placed alongside each other below, not as a complete survey but as a cross-section in which the mechanism becomes legible.

§ 04 · A dwelling built on its meter

The optimisation axis for greening the housing stock has been clear since 2013. A dwelling is judged by its energy performance. First through the EPC, from 2021 through the three BENG indicators (energy demand, primary fossil-fuel use, share of renewable energy), and in parallel through the nul-op-de-meter (NoM) standard whenever the Energieprestatievergoeding applies. For housing associations the NoM renovation is financially attractive because the Allowance can be billed as part of the housing costs once the performance is guaranteed; that is the heart of Stroomversnelling’s own formulation: *Net-zero-on-the-meter is inseparably bound to performance guarantees.*⁸ The performance is, throughout, and exclusively, the energetic performance.

What became apparent in the years following the first major NoM projects is that the concept house encompasses more than its energy balance. TNO published in May 2025 a release placing indoor air quality on the agenda as a health factor; in the same month RIVM publications appeared on the increasingly critical role of ventilation in ever more tightly built homes.⁹ The order of events is no coincidence. The energy savings made possible by air-tightening and sealed building envelopes require active mechanical ventilation as compensation. When that ventilation is poorly commissioned, poorly maintained, or switched down by occupants because of noise, indoor climates emerge in which CO₂ levels rise above 1,200 ppm, mould settles on cold bridges, and overheating in summer becomes a problem of its own.

In this way several expelled variables appear simultaneously along the production line of the NoM dwelling. Indoor air quality, because measuring it is structurally absent from the completion certificate. The acoustics of the installation, because noise from heat pump and MVHR unit is no part of the energy performance. The comfort range, because the bandwidth of comfortable indoor temperature is not the same as the bandwidth within which the meter reads zero. Repairability, because some installation components fail within five years where the building itself assumes a service life of fifteen to twenty. And one further, less easily named variable: the feeling of a house that fits the person living in it. That is not a variable that lets itself be captured by BENG indicators, and that is precisely why it no longer features in the operational description of the dwelling.

Empirically the pattern is well documented. Residents of NoM projects in Drenthe, Tilburg and Nijmegen, among others, have spoken in the press and on consumer programmes; the Woonbond, the Dutch tenants’ association, has received complaints across a series of projects about mould and comfort.¹⁰ WoON 2024 reports that around thirty per cent of rental dwellings suffer from damp and mould; CBS figures from 2024 indicate twenty-nine per cent in the housing-association stock.¹¹ The causal link between specifically the NoM renovation and specifically the incidence of mould has not been unambiguously established in public databases, and that caveat belongs here explicitly. What is mechanistically clear is how the link runs: air-tightening plus an incompletely functioning ventilation system plus cold bridges plus deviating user behaviour produces, in any combination, mould and damp risk, and the NoM renovation strengthens each of those factors at once.

What becomes visible in this dossier is how the auditability cascade operates that Michael Power has described.¹² Energy performance is auditable. Indoor air quality is so only in part, and then only through additional measurements that are not part of the standard handover. The resident’s judgement is not auditable at all, or only through follow-up research that is, after three or five years, sparingly conducted. What is auditable is called truth. What is not auditable is called ungovernable, and therefore waste. The inspector signs off. The resident’s story appears in the local paper.

In the terms of De Caluwé and Vermaak’s change colours, this is a blue intervention without any noteworthy white counterweight. Blue plans, calculates, designs, monitors. White, in this case, would have said that the dwelling will have to prove itself in use, that the resident has a meaningful voice in what counts as a well-renovated home, that the design must leave room for variation between people and seasons. The first generation of the NoM standard does not allow such considerations; it builds on the assumption that the house will fit itself to the meter, rather than the other way round.¹³ The NoM dossier connects to the housing-market dossiers in Series III Nº 02 on the congealed outcome of supply and in Series II Nº 02 on silent expropriation; where those papers investigate the ownership and preference layers, the present paper isolates the measurement axis that allows a dwelling to satisfy its meter while no longer satisfying its inhabitant.

§ 05 · A neighbourhood shop as vanished encounter

A second dossier. Picnic, founded in 2015, operates by a transparent operational principle. Groceries are gathered in distribution centres by employees who are held to a pick rate; the order is then routed via an algorithm known in the literature as VROOM to the delivery district, with twenty-minute time windows and route density as the dominant optimisation goal.¹⁴ In Utrecht and Rotterdam the company itself reports efficiency gains of sixteen to seventeen per cent thanks to the algorithm. On the warehouse floor a digital scoreboard shows each employee’s pick rate; nine crosses for slowness, illness or lateness trigger a meeting; pay for agency staff sits below the regular supermarket collective agreement; surcharges for irregular hours are absent from Picnic’s own collective agreement.¹⁵ In February 2024, the Utrecht district court ruled that Picnic, Getir, Flink and Gorillas all fall under the supermarket collective agreement and must pay surcharges retroactively.¹⁶

What does Picnic optimise? Pick rate per hour in the distribution centre, stops per route in delivery, market share in urban districts. In the Netherlands that market share stood at approximately 1.9 per cent in 2024 (Circana); group revenue including the German and French operations came to around 1.5 billion euro, with a net loss in the Netherlands of approximately 65 million euro; supermarket chain Edeka holds a 32 per cent stake in the company.¹ The figures are relevant for the proportions of the dossier: an undertaking that has been continuously loss-making since 2015 and that nevertheless grows structurally in a regulated European market does so on the basis of a choice architecture in which a few variables are maximised while others remain outside the frame. What does the system expel? In the first ring: working conditions, components of pay, the negotiability of workload. In the second ring: the neighbourhood function of doing the shopping.

That second ring is harder to make tangible, and it is right to mark this explicitly. Locatus reports in 2024 that the number of neighbourhood shops in the Netherlands has shrunk while the number of supermarkets has grown, and that market concentration at the top exceeds seventy per cent.¹⁷ Since 2000, more than twenty supermarket chains have disappeared from the Dutch market through consolidation; between 2012 and 2023 the number of small supermarkets fell from 1,052 to 622. These figures are not attributable to Picnic alone; the decline of the neighbourhood shop has a long prehistory. But the optimisation axis Picnic has chosen contributes in its own way to a tendency that was already under pressure.

For the sociological weight of that tendency, one can draw on the work of Talja Blokland and Ruth Soenen on public familiarity, and on SCP research on social cohesion in urban neighbourhoods.¹⁸ Blokland’s claim, supported by ethnographic work in Rotterdam among other places, is that social cohesion in urban neighbourhoods does not arise primarily through deep relationships but through public familiarity: recognising, greeting and knowing-by-sight the people one encounters on the street, in the shop, on the tram. Soenen adds the notion of small encounters, the brief interactions in public places in which social information is exchanged without a conversation being explicitly opened. Both authors put considerable weight on the taken-for-grantedness of the place in which those encounters occur. A neighbourhood shop, or a nearby supermarket, is such a place. A Picnic crate delivered to the door is not.

The claim that home delivery causally weakens social cohesion has not been directly tested in the literature, and the longer essay does well to note this honestly. What can be said is that the decision to build the operation of a new kind of supermarket around pick rate and stop density, and not around, for example, neighbourhood-impact indicators, is itself already a quiet value-statement. The neighbourhood function is not contested. It is not counted. That is the optimisation asymmetry: the variable that is not counted remains outside the operation, and it then surfaces as an externality elsewhere in society, where social workers, municipal policy advisers and neighbourhood councils may pick it up without being able to undo the cause.

The speech-anchor in this dossier is striking. When criticism is levelled at the workload in the distribution centre, Picnic responds with nonsense and scaremongering.¹⁹ When the supermarket collective agreement is declared applicable, the co-founder calls that agreement outdated and states that it belongs in a museum. The frame is consistent: speed and innovation against the old. What is hard to say without a heavy price is that a supermarket collective agreement is not a relic but the cumulative settlement of what earlier generations had agreed about what shopping-as-work is worthy of, and that pushing it aside under the frame of innovation is a particular form of what this paper names: a variable expelled because it has no place in the operational description of the new system. The Picnic dossier appears in The Congealed Outcome as Preference (Nº 02) as the principal illustration of the rhetorical mechanism by which the outcome of a supply-side decision is presented as consumer preference; in the present paper the dossier serves to elaborate what precedes that rhetoric, namely the choice to build the operation of a new kind of supermarket around a single maximised variable.

§ 06 · A civil-service post without memory

A third dossier, into the heart of central government. The Algemene Bestuursdienst (ABD, Dutch Senior Public Service), established in 1995 to foster mobility, broad deployability and professionalisation of senior civil-service posts, comprises in 2024 around 1,900 management positions at salary scale fifteen and above. The Dutch civil-service collective agreement stipulates a minimum tenure of four years for those posts and, for top management, a maximum of seven years as a rule; in 2024 the average tenure according to the ABD annual report stood at four years and eight months, with 358 fresh starts on ABD posts in that year.²⁰

What does the ABD system optimise? Mobility as a virtue. Broad deployability as a quality. The avoidance of senior officials sitting too long. A career path on which a director can rotate through three ministries within six years. What does the system expel? Dossier knowledge. Networks within the field. The slow accumulation of understanding for the peculiarities of a specific delivery chain.

The parliamentary inquiries of the past decade — the childcare-allowance affair, the Groningen earthquake dossier, fraud detection at the Dutch Tax Administration — have repeatedly pointed in their reports to civil-service discontinuity as one of the factors that deepened the harm. When in a delivery chain that is to play out over twenty years a director is replaced after four years by someone without prior dossier knowledge, and that successor in turn rotates onwards after another four, an institutional loss of memory arises that translates over time into concrete damage. Werk aan Uitvoering (2020) and the writings of Tjeenk Willink from the same period²¹ have named this. Under the Schoof government, a first step towards reform of the ABD was announced in March 2025, with greater emphasis on leadership qualities, substantive expertise, knowledge of practice, awareness of the rule of law and a direct connection with citizens.²² That is an implicit acknowledgement of the optimisation asymmetry. The announcement is not the evaluation of a successful correction; it is the recognition that the pattern is at work and that it requires counterweight.

A methodological note belongs here. The ABD annual report records aggregates, not function-specific cohorts. A claim such as the average tenure on a Secretary-General post within a given ministry has, over thirty years, fallen from X to Y years cannot be reconstructed from public sources; that would require ministry-by-ministry, cohort-by-cohort enquiry. What can be established from public sources is that the minimum tenure in the collective agreement is low, the average tenure narrowly above that minimum, and mobility-as-goal independently celebrated in the annual reports. Dossier knowledge does not appear anywhere as a KPI. The civil-service presence at earlier cabinet crises, on related dossiers from a decade ago, in operational knowledge of what actually happens in the delivery chain — none of this appears as a variable that the system optimises. It has become waste of a process that was aimed at something else.

In an interim assignment at a centre municipality (centrumgemeente, the regional lead municipality for a designated policy domain) in the eastern Netherlands, in the social domain, I saw some years ago how this pattern mirrors itself at municipal level. The director of the social domain had moved on to a neighbouring region after two and a half years. His successor took up a dossier whose main outlines were new to him, while the delivery partners and the local social teams had been wrestling with the same bottlenecks for years. The files had been carefully handed over. The names, the telephone numbers, the relational history with the providers, the why behind the choices made two years earlier — all of that had to be rebuilt from scratch. In the two months during which that rebuilding took place, decisions on a complex care pathway stagnated. This is no anecdote against mobility. It is an observation that mobility as the sole optimisation axis carries costs that show up in no budget line. That the ABD system returns in Series I (Dissociated Organisations) as an example of an institutional measurement logic that has detached itself from the primary work, and in the Series II paper on the childcare-allowance affair as the reverberation for citizens, illustrates how this same dossier is legible across several layers. The present paper isolates the measurement axis itself.

§ 07 · A care relationship as side-issue

A fourth dossier. Since 2015 youth care in the Netherlands has been decentralised to the municipalities, and since then municipalities have favoured an arrangement that has come to be known in procurement practice as Open House. The municipality establishes terms, a tariff and a quality framework; any provider that meets the conditions may contract under those terms. The arrangement is attractive because it suggests freedom of choice for the client and because it side-steps the obligation to procure under EU rules; at the same time it is fragile, because it contains no incentive whatever to limit the number of providers or to select on quality.

The figures are known. In ten years the number of youth-care providers grew, according to Chamber of Commerce data, by more than seven thousand.²³ One Groningen cluster went from thirty to two hundred and twenty providers in seven years. Arnhem saw the number of youth-care providers rise from three hundred to nine hundred between 2015 and 2020, with the consequence that eighty-six per cent of new clients ended up at smaller providers.²⁴ In Maastricht, according to internal documents from 2020, eighty per cent of the budget went to ten providers, while a portion of the contracted providers saw no client at all. The CPB observed in 2021 that under light selection (Open House) a municipality contracted 23.9 ambulatory providers per thousand children, against 2.9 under strict selection; the CPB added that a difference in care volume between the two systems could not be demonstrated.²⁵ In other words: the fragmentation produced no demonstrable advantage in volume or accessibility, but it did produce an explosion in administrative load and in the fragility of the system.

What does this arrangement optimise? Formal lawfulness of the procurement process. Avoidance of the EU procurement obligation. Theoretical freedom of choice for the client. Cost control per trajectory, at least at the design stage. What does the system expel? The continuity of the care relationship. The Vergeten Kind Foundation documented in 2021 that children in residential-group arrangements had had on average sixty-four point six different carers in their lives; that almost three quarters of them found the number too high; and that for a quarter of that group the situation worsened as a result of the changes.²⁶ The Netherlands Youth Institute reports that three out of five children with multiple problems change carer against their will.²⁷ The therapeutic literature has known for decades that continuity of the therapeutic relationship is one of the strongest predictors of treatment outcome; for the target group of children in foster care, children with attachment disorders, young people with complex trauma, that continuity is not merely a precondition for effective treatment, it is part of the treatment itself.

The pattern repeats itself in the waiting times. The Vergeten Kind Foundation documented in 2021 that a child in distress waited on average forty-four weeks for help; eighty-one per cent of the cases tracked encountered a waiting period; fifty-six per cent of children removed from home ended up in a less suitable placement than had been indicated.²⁸ At the same time, Andersson Elffers Felix calculated that Dutch youth care had a structural shortfall of 1.6 to 1.8 billion euro a year.²⁹

Hugo de Jonge, in 2018 the responsible minister, formulated the self-criticism in sharp terms. Open House can lead to enormous fragmentation of supply, and under this arrangement the municipality does not take the reins, but places them entirely in the hands of individual care-seekers and providers. And further: *care and youth services are not a market, and certainly not a European market.*³⁰ The minister in question is the same one who later signed up to the Youth Reform Agenda, with as instruments a Decree on Real Tariffs (AMvB Reële Prijzen) that came into force on 1 July 2024 and an expert commission under Tamara van Ark which delivered its interim advice in January 2025.³¹ These are corrections that acknowledge the pattern. Whether they break the pattern cannot be established at the time of writing, and that should be explicitly noted here.

What becomes visible in this dossier is how optimisation asymmetry is not solely an internal corporate failure but can travel between layers of government and between sectors. The municipality that procures on cost expels continuity. The provider that optimises for tariff volume expels continuity equally. The child who has to undergo the change carries the expelled variable in his own developmental history. Lipsky’s street-level bureaucrat — in this case the youth carer — stands between the aggregation of the procurement system and the individual relationship in which the work actually takes place, and the aggregation almost always wins, because only the aggregation is visible to those who judge the carer.³² The Open House arrangement returns in Series III Nº 05 as the principal illustration of the pattern in which the problem-causer simultaneously becomes the solution-provider: a procurement mechanism that produces fragmentation and subsequently creates a market for combating that fragmentation. In the present paper the dossier serves to illustrate the optimisation asymmetry that precedes that rhetorical mechanism.

§ 08 · A consulting room without context

A fifth dossier, briefer than the preceding ones because the structure has by now become clear. The Dutch DBC system (Diagnose-Behandel-Combinatie, Diagnosis-Treatment-Combination), introduced in 2005 in specialist medical care and since grown to roughly 4,500 unique care products, registers the treatment of a patient as a combination of diagnosis and care product, with a standard duration of at most one hundred and twenty days.³³ The tariff is set by the Nederlandse Zorgautoriteit (Dutch Healthcare Authority) on the basis of average care delivered and average costs. In general practice a comparable system has been introduced, with different product definitions but the same underlying logic.

What does the system optimise? Processing time. Production norm. Billable activity. What does the system expel? Time-with-the-patient that cannot be billed. Contact with other carers about a complex case. Reflection. The conversation with parents during a paediatric consultation. The silent minute in which a patient gathers themselves to say something painful.

It is not a critique of evidence-based medicine as such to observe that the DBC system has translated evidence-based medicine into something that might more accurately be called registration-based. The Council for Public Health and Society (RVS) noted in 2017 that evidence-based medicine in Dutch healthcare has been interpreted too narrowly, that the quest for evidence leads to the filtering out of context, and that good care requires a permanent dialogue with the context of patients as input.³⁴ The Council introduced the term context-based medicine as a corrective. The report led to vigorous debate within the medical profession, with counter-criticism from, among others, the general-practice guidelines tradition. That the debate was conducted so fiercely is itself an indication that the Council had touched a sensitive nerve: the burden of proof for more time with the patient had come to rest with the doctor, while it ought to have rested with the system that had made less time into operational norm.

In an interim assignment in a municipality of around one hundred thousand inhabitants, some years ago, I heard a general practitioner say at a neighbourhood meeting that he had seven minutes per consultation, of which two minutes went to registration, ICPC coding and validation of insurer contracts. The screen sits between me and the patient, he said. The patient is telling me about her daughter, and I am looking at the screen. That is not a complaint about digitisation. That is the precisely correct articulation of what optimisation asymmetry looks like in the consulting room. The variable that mattered at that moment — attention — was not billable. The variable that had been optimised — time per consultation — was structured down to the second. That same DBC system functions in Series III Nº 05 as the material environment in which private equity can enter the chain through product definitions tuned to return optimisation; in the present paper it is the measurement axis that makes the expulsion in the consulting room possible.

§ 09 · Five domains, one regularity

The five dossiers together display the same structure. A variable that is in itself meaningful is bound to operationalisation, to a norm, and to consequences. The other variables that, in the previous system, took part in defining the work remain formally namable but disappear from the operational description. They become waste. The becoming-waste is silent. No one explicitly declares the new definitional frame narrower than the old; the narrowing accomplishes itself through the structure.

The pattern has long been diagnosed internationally. Gwyn Bevan and Christopher Hood described in 2006 how the British NHS in the early years of the 2000s operated under a target regime that they called targets and terror, after Soviet precedent.³⁵ Patients were held in ambulances outside the Accident & Emergency department so that the four-hour clock would not start. Patients were re-classified into non-targeted categories. What’s measured is what matters, Bevan and Hood wrote, with the sharper variant: hit the target and miss the point.

In the United States, No Child Left Behind (2002 to 2015) produced a comparable dossier. Diane Ravitch, originally an advocate, described the regime in 2010 as a timetable for the destruction of American public education; the Atlanta cheating scandal, with one hundred and seventy-eight teachers and administrators charged for correcting student answers, illustrated Campbell’s law in its purest form.³⁶

In the New York Police Department, John Eterno and Eli Silverman documented how the CompStat system, originally an instrument for district analysis and operational steering, was under Bratton and his successors converted into a tool through which precinct commanders were judged on crime numbers and through which systematic downgrading of offences took place to drive the figures down.³⁷ Seventy-eight per cent of CompStat-era commanders indicated in a survey that post-incident adjustments to crime reports were, to some degree, unethical.

In China, a series of studies including those by Frank Zhang and colleagues has documented how provincial and prefectural GDP figures are systematically inflated when official growth targets approach, particularly at the final promotion moment of the relevant party official.³⁸ Alec Nove documented for the Soviet economy the classic specimens of the pattern: nails by weight, where the nail became too heavy; glass by surface area, where the glass became too thin; paper by weight, where the paper became too thick.³⁹

None of these parallels constitutes a direct equivalence with the Dutch situation. Yet they show how robust the regularity is. In authoritarian and democratic political orders, in command economies and market-rich welfare states, in healthcare, education, policing and industry: as soon as a KPI carries consequences, distortion appears, and the un-measured dimension becomes waste. The Netherlands is no exception, except that the pattern is here borne in milder form and with less cynicism, and is for that very reason harder to surface.

§ 10 · The cognitive structure

Why is the pattern so persistent? Four elements together account for its persistence.

First, accountability cycles compel quantification, and quantification requires operationalisation. A budget, an annual report, a court of audit report, a healthcare-authority decision, a building-energy assessment: in every accountability cycle, reality must be returned in numbers. Without numbers there is no accountability. Numbers require operationalisation: the concept must be captured in a measurable proxy. Operationalisation narrows the concept. Quality becomes score on a questionnaire. Comfort becomes indoor temperature. Continuity becomes number of carers per year. That is no distortion of the concept; that is how operationalisation works. The distortion arises as soon as the proxy replaces the concept itself. When in the spreadsheet quality equals score on questionnaire, and only the spreadsheet is visible in the decision process, the score has become quality. Goodhart in his most philosophical form.

Second, Power’s auditability cascade. What is not auditable is called ungovernable; what is auditable is called truth. The accountant, the audit office, the inspectorate, the healthcare authority: they ask for verifiable data. Practices are reorganised to be verifiable, not because the reorganisation improves the primary work. Power calls this making things auditable. The ABD annual report records mobility figures because they are auditable. It does not record dossier-knowledge depth because that is not auditable. The completion certificate of the NoM renovation records energy performance because it is auditable. It does not record resident satisfaction because that is not auditable, at least not within the standard pathway.

Third, the optimisation spiral. Once a KPI is in place, every improvement on it becomes a rationalising argument for further narrowing. Pick rate has risen from X to Y, therefore operational efficiency has improved. That argument itself makes it politically difficult to roll back to a broader concept. Wendy Espeland and Michael Sauder have documented this for university rankings: once an institution joins, it costs more to step out than to adapt further.⁴⁰ The spiral is not exclusively technical. It is moral: anyone resisting the optimisation appears to resist the gain previously achieved, and bears the burden of proof for that resistance.

That brings the fourth element. Whoever criticises the KPI carries the burden of proof. The doctor who says I want more time with my patient has to demonstrate that the time adds value, while the DBC system ought to have demonstrated that the narrowing came without loss. The civil servant who says I prefer to stay on this dossier has to explain why he is not rotating onwards, while the system ought to have explained why continuity has become a subordinate virtue. The housing-association employee who says this NoM renovation is not sound has to provide research, while the handover process should have proven that the dwelling is a dwelling. This is what I have called in an earlier paper the moral shadow: the optimisation tints the prohibition on criticism in the colours of argument. Aiki work in the strict sense, the work of moving with the energy of the system, becomes, without an ethical footing, work in service of the pattern; in The Sincere Voice (Series III, Nº 01) this is set out precisely through the case of the housing-construction director.⁴¹

In the terms of the change colours, the pattern is prototypically blue. It designs, plans, measures, sets norms. That in itself is no pathology; many public processes call for blue intervention. The pathology arises when blue becomes monochromatic, that is, when no other colour offers counterweight. White, which gives space to self-organisation and emergence, cannot come to the table because self-organisation is not auditable. Red, which gives space to motivation and people, cannot come to the table because motivation is not quantifiable in the cycle. Green, which gives space to learning, cannot come to the table because learning requires a time horizon longer than the budget cycle. What remains is a blue monoculture in which the pattern is unavoidable.

§ 11 · Diagnostic questions

For anyone seeking to recognise the pattern in themselves and in their organisation, the structure leads to five questions.

Materially: which variable in my organisation is measured and reported in the monthly meeting with my line manager or in the annual report, and which essential variable is not in that conversation? Can I name aloud what I therefore have to systematically expel? When an organisation answers the question well, the second half of the answer will sound uncomfortable, because it names what does not exist in the formal accountability.

Linguistically: what words does the organisation use when someone questions the KPI? Subjective, un-measurable, anecdotal, old-fashioned? And what words for those who embrace the KPI? Data-driven, evidence-based, transparent, professional? The asymmetric vocabulary betrays the moral shadow. One side comes equipped with legitimising language, the other with delegitimising language, while the empirical relationship between the two is rarely so straightforward.

Temporally: if I look back five years from now on the current intervention — not what the meter reads on the day of completion, but what still stands when no one is thinking about the intervention any longer — which variables will then turn out to have been cut away? This is the embedding test (in Dutch borging, the long-term anchoring of an intervention in structure, persons and processes such that it survives without continuing attention) from chapter nine of De Richting van de Beweging (The Direction of the Movement): success is not measured on the day of departure, but in what remains in place after the handover.⁴² The optimisation asymmetry almost always reveals itself on the embedding test, because what was expelled at handover will, in time, come back through the door of the work.

Socially: which voice in the organisation speaks about the expelled variables? Is that voice at the table or in the corridor? Is it heard or labelled as not constructive? Does it sit in the construction meeting or in the cleaning cupboard? In healthy organisations, someone in a recognisable position speaks about the expelled variable, and that voice is taken into account in decisions. In organisations with deeply embedded optimisation asymmetry, the voice does not belong at the table, or it sits there ironically.

Reflexively: do I myself carry the KPI with conviction? Have I ever publicly committed myself to a figure of which I knew, deep down, that it narrowed reality? This is the direct connection to The Sincere Voice. The pattern derives its institutional force not from the cynics on the outside but from the sincere people on the inside. Anyone who recognises themselves within the pattern does not thereby have an exit, but does have a starting point.

A sixth question, following from Power’s work: which parts of my work are auditable, and which are not? Could I sacrifice the non-auditable parts without the system noticing? Whenever the answer is yes, and whenever the answer is yes for essential parts of the primary work, the optimisation asymmetry is deeply built in.

§ 12 · What breaks the pattern

Breaking the pattern requires structural counterweight, not additional measurement. More measurement entrenches the logic; counterweight does not. Five lines of correction are visible in practices that have managed to turn or contain the pattern.

First, designing from multiple goals at once. The methodological counterpart to single-objective optimisation is multi-criteria decision analysis: every decision explicitly weighs all relevant dimensions, with named Pareto trade-offs.⁴³ For housing, this would mean that alongside BENG indicators, indoor air quality, noise, repairability, comfortable temperature range and resident satisfaction figure in the assessment framework — not as appendix but as equal criterion. For youth care, it would mean that continuity of carer, waiting time and substantive outcome measures sit alongside cost in the procurement cycle, with explicit acknowledgement of the trade-off. The method is technically available; its adoption is a political choice.

Second, the restoration of professional judgement. Onora O’Neill formulated in her 2002 Reith Lectures the notion of intelligent accountability, a form of accountability that gives space to context, judgement and reasons, rather than performance indicators that demand only the form.⁴⁴ We may undermine professional performance and standards in public life by excessive regulation, O’Neill wrote, and we may condone and even encourage deception in our zeal for transparency. Atul Gawande’s work on checklists and Iona Heath’s work on general practice point in the same direction: instruments that support professional judgement work; instruments that replace judgement erode the capacity for judgement itself.

Third, institutional counter-examples that have been explicitly designed to avoid the pattern. Buurtzorg, founded in 2006 by Jos de Blok, organises home care in self-managing teams of at most twelve nurses, with minimal administration and with the explicit ambition of placing humanity over bureaucracy.⁴⁵ In youth care, the Dutch municipality of Veendam has reshaped its arrangement around a Youth Expertise Point, with universal provisions free of indication and with a support worker present one day a week at every general practice and primary school; the municipality reports that waiting lists have disappeared and costs have fallen.⁴⁶ The Finnish education experience, in which national high-stakes testing does not exist and trust in the teacher as professional is central, is much discussed in the literature; its transferability to the Netherlands is limited, but the design choice is instructive. The passive house, as an alternative to the strict BENG norm, adds a comfort range to the energy performance and so makes the dwelling less vulnerable to the expulsion of those other housing variables. None of these examples is free of conflict, and none is a blueprint. Their common feature is that they explicitly add to the measurement axis rather than replace it.

Fourth, the audit of the audit, in Power’s own proposal. Who audits the auditor? What costs does the audit itself generate? Which practices have been cut away in order to become auditable? A Dutch version of that audit-of-the-audit could, for example, register how much time general practitioners spend on administration versus contact, as the NIVEL research institute does, and incorporate those meta-figures into the budget cycle.⁴⁷ The effect would be that the pattern makes itself visible and can be weighed in the political conversation.

Fifth, long-form journalism and narrative inquiry as counterweight to numerical reporting. The Follow the Money investigation into Open House youth care in 2020, jointly with twelve regional newspapers, formed the empirical basis for De Jonge’s self-correction. The BNNVARA Kassa format documents NoM and mould problems in a way that policy documents cannot. Pointer and Investico have, in a series of programmes, made the implementation reality of benefits administration and youth care visible.⁴⁸ This is no sentimental story-versus-figure. It is a methodological correction. Numerical reporting delivers what it delivers; narrative inquiry delivers what numerical reporting cannot; only in the combination does an account emerge in which the expelled variable comes back into view.

A statutory limit on KPI management exists in the Netherlands only to a limited degree. The Decree on Real Tariffs in youth care (1 July 2024) is a floor. The removal of the EMVI criterion for youth and Wmo procurement in 2022 was a correction. A general statute that would limit the binding of KPIs to reward in the public sector does not exist, and it is conceivable that a later generation will have to design one.

§ 13 · Connection

The paper does not stand alone. With The Sincere Voice (Series III, Nº 01) it shares the observation that the pattern derives its institutional force from those who carry it with conviction, not from those who apply it cynically.⁴⁹ With The Congealed Outcome as Preference (Series III, Nº 02) it shares the structure in which an outcome caused by earlier design choices is subsequently presented as a manifested preference; the KPI outcome of optimisation is read back narratively as what the market wanted or what consumers wanted. Where Nº 02 elaborates that rhetorical mechanism as a pattern in its own right and traces it through a series of Dutch dossiers, the present paper addresses the operational condition that makes the pattern possible: the single-objective optimisation that produces the outcome in the first place.

With Series I, Dissociated Organisations the paper shares the claim that the measurement comes to reside in a dissociated subsystem that develops independently of the primary work process. The KPI department speaks its own language, meets in its own rhythm, uses its own data, and grows detached from what happens in the consulting room, the warehouse, the building site. That is not a by-product but the operational core of what the first series named institutional dissociation.⁵⁰

With Series II, Doorwerking (Reverberations) the paper shares the observation that optimisation steers the reverberation onto the measurable citizen. Whoever is visible in the system becomes more visible; whoever is illegible falls through the cracks or is wrongly caught. That is not only the story of the childcare-allowance affair; it is a continuing observation across every form of public policy in which the client or the citizen appears primarily as a data point.

With De Richting van de Beweging (manuscript in preparation), the paper shares the core perspective on embedding. Chapter nine of the manuscript treats embedding as the first KPI of interim work, with the claim that success is measured in what still stands at the moment when no one is thinking about the intervention any longer. For the pattern described here, embedding is the most reliable diagnostic touchstone there is. What depends on a KPI binding that looks impressive at the moment of reporting, and which falters or turns into its opposite after the executive’s departure or after a change of cabinet, is not embedded. What is embedded has carried the expelled variable along in the design; what is not embedded has driven it out.

A final connection belongs with the pamphlet The Discriminating Eye, available at nourishment.houseofviridian.org since April 2026.⁵¹ The pamphlet works out a five-stage trajectory of brands (workshop, reputation, scale, conglomerate, hieroglyph) and formulates ten commandments, several of which are directly applicable to the pattern of this paper. The sixth commandment, trust the eye over the metric, and the eighth, measure what makes the work, not what justifies it, run parallel to the structure laid out here. The pamphlet is no footnote to this paper. It is a toolkit, parts of which apply to the pattern.

§ 14 · Closing

What this paper has named is not a plea against measurement. A public sector without numbers is no public sector but a collection of personal preferences. Accountability requires numbers; numbers require operationalisation; operationalisation narrows the concept. There is no path around that, and it would be a mistake to want to find one.

What has been named is something more specific. When one variable in a multi-valued system is bound to reward or sanction, and when no institutional counterweight is at the table, the system moves towards a corner of its design space in which the un-measured dimensions are in free fall. The variable that was meant to safeguard care, housing, implementation, neighbourhood or work becomes, in its very success, the instrument that consumes its original object. When a measure becomes a target, it ceases to be a good measure.

The ground on which the pattern can be broken is not methodological and not technical. It is institutional. It requires that a voice for the expelled variable sits at the table, with authority, with mandate, with budget, with the ability to place the KPI in its context without having to treat the KPI as enemy. What such a counter-voice needs is a design that treats it not as luxury but as constitutive. What current Dutch implementation shows is that the design is locally possible and that the structural variant has not yet been built.

The dwelling in Drenthe is still standing. The meter reads zero. In the kitchen someone draws the blanket around herself. What this paper diagnoses is not that the meter is wrong. It diagnoses that the system that invented the meter has forgotten the woman who lives there, and that the forgetting was no accident.

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.

Footnotes

Colophon

About the author Jacob Huibers is an interim manager, author and adviser in the Dutch public sector, with assignments in the social domain, the physical domain, regional partnerships and administrative recovery in municipalities of fifty thousand to two hundred and fifty thousand inhabitants. He is the author of De Richting van de Beweging: Interim-Management in de Publieke Sector (manuscript in preparation) and of the corpus Limbic Literacy, Allemaal Ontheemd and Decline and Revival, all published under House of Viridian.

About the series Series III is the third Statecraft series of House of Viridian. It teaches public officials and citizens to recognise five form-patterns of cognitive distortion that remain invisible in the soft layer of policy but become legible in hard materiality (housing, food, objects, infrastructure, transmission). The five form-patterns are carried by one meta-pattern and closed by a synthesis. Materiality is less mendacious than documents because it does not change quickly enough to carry the illusion of exogeneity convincingly. Series III is not a third diagnostic layer alongside the first two series. Series I names the mechanism of institutional dissociation, Series II describes the reverberation for citizens, and Series III teaches the eye how to see.

This paper is the fourth in Series III. The first paper, The Sincere Voice, opened the series and addresses the meta-pattern that clothes the other five: the most powerful carriers of systemic inversion are the sincere people inside the system, not the cynics outside it. The second paper, The Congealed Outcome as Manifested Preference, describes how the outcome of earlier design choices is then presented as a consumer preference. The third paper, Word Continuity that Masks the Material Rupture, addresses the pattern in which the word house or care still does the work that the current content can no longer carry. The present paper, The Optimisation Asymmetry, describes how one measurable variable cuts away the others and what becomes of what is not counted. The fifth paper, The Problem-Causer as Solution-Provider (Nº 05), describes how parties that helped produce the problem subsequently position themselves as solution-providers and so entrench a diachronic dependency. The sixth paper, Form Laundering (Nº 06), describes how institutions retain the outward features of functions they can no longer deliver, and how this is not deception but a design choice.

Place in the series Series III Nº 01: De oprechte stem (The Sincere Voice) — meta-pattern. Series III Nº 02: De gestolde uitkomst als gemanifesteerde voorkeur (The Congealed Outcome as Manifested Preference) — Pattern 1. Series III Nº 03: De woordcontinuïteit die de materiële breuk maskeert (Word Continuity that Masks the Material Rupture) — Pattern 2. Series III Nº 04: De optimalisatie-asymmetrie (The Optimisation Asymmetry) — Pattern 3. Series III Nº 05: De probleemveroorzaker als oplossingsleverancier (The Problem-Causer as Solution-Provider) — Pattern 4. Series III Nº 06: De vorm-laundering (Form-Laundering) — Pattern 5. Series III Nº 07: Synthese (Synthesis).

Publisher HOUSE OF VIRIDIAN OÜ Tallinn · Lisbon

Contact jacob@statecraft.nl statecraft.nl

Series: STATECRAFT SERIES · SERIES III Nº 04

Footnotes

¹ Stroomversnelling, Eerste versie Energieprestatie Monitoring Norm beschikbaar, press release 2018, and Monitoren van energieprestaties biedt bouwbedrijf kansen, Energielinq 2019. For the financing structure via the Energieprestatievergoeding, see Aedes and Stroomversnelling, Handreiking EPV, updated 2022. For the construction and handover practice: ISC, Nul op de meter vooralsnog een zorgenkindje, sector article 2019.

² The quotation has been taken from a series of regional press reports and consumer programme features on Net-Zero-on-the-Meter renovations in Drenthe, Tilburg and Nijmegen, in the period 2017–2019. For the specific Drenthe case, see consumer-programme reporting on housing-association projects in the region of Borger-Odoorn and surroundings. The quotation here functions as exemplar and is not tied to a still-living, identifiable individual; comparable formulations have been recorded across multiple reports.

³ M.H. Moore, Creating Public Value: Strategic Management in Government, Harvard University Press, 1995. For the elaboration of the Strategic Triangle within the Statecraft corpus, see J. Huibers, Statecraft in the Interregnum, April 2026, Statecraft Series Nº 04, and De Richting van de Beweging: Interim-Management in de Publieke Sector, manuscript in preparation, chapter 3.

⁴ C.A.E. Goodhart, Problems of Monetary Management: The U.K. Experience, Reserve Bank of Australia conference 1975, reprinted in C.A.E. Goodhart, Monetary Theory and Practice: The U.K. Experience, Macmillan, 1984. The canonical formulation runs: any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes. For an analysis of the genesis and interpretation of Goodhart’s law, see K. Chrystal and P. Mizen, Goodhart’s Law: Its Origins, Meaning and Implications for Monetary Policy, in Mizen (ed.), Central Banking, Monetary Theory and Practice, Edward Elgar, 2003.

⁵ D.T. Campbell, Assessing the Impact of Planned Social Change, in Evaluation and Program Planning 2(1), 1979, pp. 67–90. Earlier formulations by Campbell from 1969 onwards; see also J. Rodamar, There ought to be a law! Campbell versus Goodhart, in Significance 15(6), 2018.

⁶ M. Strathern, ‘Improving Ratings’: Audit in the British University System, in European Review 5(3), 1997, pp. 305–321; the cited formulation on p. 308.

⁷ W.N. Espeland and M.L. Stevens, Commensuration as a Social Process, in Annual Review of Sociology 24, 1998, pp. 313–343, and A Sociology of Quantification, in European Journal of Sociology 49(3), 2008, pp. 401–436. See also W.N. Espeland and M. Sauder, Engines of Anxiety: Academic Rankings, Reputation, and Accountability, Russell Sage Foundation, 2016. Further: J.C. Scott, Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed, Yale University Press, 1998. M. Power, The Audit Society: Rituals of Verification, Oxford University Press, 1997. J.Z. Muller, The Tyranny of Metrics, Princeton University Press, 2018. For mechanical objectivity as a symptom of distrust of professional judgement: T.M. Porter, Trust in Numbers: The Pursuit of Objectivity in Science and Public Life, Princeton University Press, 1995.

⁸ Stroomversnelling, Eerste versie Energieprestatie Monitoring Norm beschikbaar, retrieved via stroomversnelling.nl. The Energieprestatie Monitoring Norm (EPMN) is explicitly designed to substantiate the level of the Energieprestatievergoeding.

⁹ TNO, Verbeter binnenluchtkwaliteit voor een betere gezondheid, press release May 2025; RIVM, Binnenmilieu-kwaliteit: ventilatie en vochtigheid, dossier on rivm.nl, retrieved 2026; RIVM, GGD-richtlijn medische milieukunde: beoordeling van ventilatie en ventilatievoorzieningen van woningen, report 609330011, 2010.

¹⁰ BNNVARA Kassa, Klachten bij nul-op-de-meterwoningen, and Schimmel en vocht zorgt in één op de vier huurwoningen voor problemen. NOS Nieuwsuur, Schrijnende situaties door schimmelwoningen, ‘groot nationaal probleem’. Woonbond, Wat kan ik doen tegen schimmel in huis?, and the Woonbond complaints register. The reporting on Tilburg, Nieuw-Buinen and Nijmegen-Meijhorst runs through these sources and through local media.

¹¹ WoON 2024, chapter on quality and suitability of dwellings; CBS, Vocht- en schimmelproblemen in Nederlandse huurwoningen, thematic publication 2024.

¹² M. Power, The Audit Society, op. cit. For the specific notion making things auditable see the central chapter on auditability and the transformation it requires of practices.

¹³ L. de Caluwé and H. Vermaak, Leren Veranderen: Een handboek voor de veranderkundige, first edition Samsom, 1999; third fully revised edition, Vakmedianet, 2019. For the elaboration of blue intervention without a white counterweight in the Statecraft corpus, see J. Huibers, De Richting van de Beweging, manuscript, chapter 8.

¹⁴ J. van Tatenhove et al., VROOM: Vehicle Routing Open-source Optimization Machine, descriptions and analyses on TTM.nl and ManagementSite.nl, 2019–2020. The algorithm combines classical vehicle routing with machine-learning components for predicting doorstep times.

¹⁵ NRC and Trouw, joint investigation Picnic in 2018, 15 December 2018; FNV, press statement following interviews with more than one hundred (former) employees, 2018–2019. For the wage comparison with the supermarket collective agreement: arbeidsvoorwaardennieuws.nl, 2019.

¹⁶ District court of Midden-Nederland, Utrecht location, ruling February 2024 in cases brought by FNV against Picnic, Getir, Flink and Gorillas. For the legal context: Arbeidsvoorwaardennieuws, Picnic en flitsbezorgers verliezen, ook voor hen geldt de supermarkt cao.

¹⁷ Locatus, Retailmarkt 2024; CBRE, Hoe ziet het supermarktlandschap er over vijf jaar uit?, market report 2023. For market concentration, see also ACM, Marktscan supermarkten, periodic publication.

¹⁸ T. Blokland, Community as Urban Practice, Polity Press, 2017; R. Soenen, Het kleine ontmoeten: over het sociale karakter van de stad, Garant, 2006; SCP, Samenleven in de toekomst, December 2024, and Samen verschillend, 2024. For the broader background, see WRR, De nieuwe verscheidenheid, 2018, and Samenleven in verscheidenheid, 2020.

¹⁹ Logistiek Profs, Picnic is het niet eens met kritiek op werkdruk: ‘Stemmingmakerij’; Emerce, Picnic: ‘De werkomgeving onveilig, de werkdruk te hoog en de beloning ondermaats’ (update). The remark on the collective agreement comes from interviews with M. Muller in business media in the period 2023–2024.

²⁰ Ministry of the Interior and Kingdom Relations, Jaarverslag Algemene Bestuursdienst 2023, and Jaarverslag Algemene Bestuursdienst 2024. For the collective-agreement provisions on tenure: cao Rijk, chapter 17.

²¹ ABDTOPConsult, Werk aan Uitvoering, 5 February 2020 (Phase 1) and 3 July 2020 (Phase 2); H. Tjeenk Willink, Groter denken, kleiner doen: een oproep, Prometheus, 2018. For the childcare-allowance affair context, see Parliamentary Inquiry Committee on the Childcare Allowance, Ongekend onrecht, 17 December 2020.

²² Government of the Netherlands, Kabinet zet eerste stappen richting hervorming Algemene Bestuursdienst (ABD), press release March 2025; ABD annual report 2024, chapter Hervormingsagenda DGABD, launched in 2024, with information to the House of Representatives in March 2025.

²³ Chamber of Commerce data presented in EenVandaag, In 10 jaar ruim 7.000 jeugdzorgaanbieders erbij, Veendam besloot in te grijpen, and confirmed in research by Follow the Money and regional newspapers.

²⁴ J. Hardij et al., Inkoopformule ‘open house’ is doping voor jeugdzorgbedrijven, Follow the Money investigation 2020, in collaboration with twelve regional newspapers.

²⁵ CPB Netherlands Bureau for Economic Policy Analysis, De gemeentelijke inkoop van jeugdzorg, CPB Notitie 2021. For the broader research line, see also N. Uenk, Inkoop sociaal domein (PPRC).

²⁶ Stichting Het Vergeten Kind, Week van Het Vergeten Kind 2020 (February 2020), for the figure of an average of 64.6 carers for children in residential-group arrangements; Stichting Het Vergeten Kind, Onderzoek: kind in nood wacht gemiddeld 44 weken op hulp (June 2021), for the waiting-time figures (44 weeks on average, 81 per cent of children had to wait, 56 per cent of those removed from home placed in a less suitable location).

²⁷ Netherlands Youth Institute, Kind wisselt vaak tegen zijn zin van hulpverlener, thematic publication 2021.

²⁸ Stichting Het Vergeten Kind, op. cit. (note 26).

²⁹ Andersson Elffers Felix, Stelsel in Groei: een onderzoek naar de financiële impact van het brede transformatietraject in de jeugdzorg, 2020. For the current situation 2024–2025, see Expert Commission Tamara van Ark, Tussentijds advies Hervormingsagenda Jeugd, 30 January 2025.

³⁰ H. de Jonge, quoted in J. Hardij et al., Inkoopformule ‘open house’ is doping voor jeugdzorgbedrijven, Follow the Money, 2020. For De Jonge’s later position: Binnenlands Bestuur, Minister wil af van Open House in jeugdzorg.

³¹ Government of the Netherlands, Kamerbrief amvb reële prijzen Jeugdwet en 2 rapporten, 22 March 2024; VNG, AMvB reële prijzen Jeugdwet sinds 1 juli van kracht, 2024; Voor de Jeugd & het Gezin, Tijdlijn Hervormingsagenda Jeugd, and Government of the Netherlands, Deskundigencommissie toetst de uitvoering van de Hervormingsagenda, 2 April 2024.

³² M. Lipsky, Street-Level Bureaucracy: Dilemmas of the Individual in Public Service, Russell Sage Foundation, anniversary edition 2010 (originally 1980).

³³ Nederlandse Zorgautoriteit (NZa), Registreren en declareren medisch-specialistische zorg, ongoing publication on nza.nl. For patient-facing explanation, see Ziekenhuis Gelderse Vallei, Wat is een DBC?, and Zorgwijzer, DBC: betekenis en uitleg systematiek.

³⁴ Raad voor Volksgezondheid en Samenleving (Council for Public Health and Society), Zonder context geen bewijs: over de illusie van evidence-based practice in de zorg, June 2017. For the debate within the medical profession: J.S. Burgers and H. van Barneveld, De dokter moet leidend blijven in het DBC-systeem, Medisch Contact, 2017; R. van der Zwet, RVS maakt stropop van evidence-based practice, Sociale Vraagstukken, 2017; reactions in Bijblijven, themed issue 2018.

³⁵ G. Bevan and C. Hood, What’s Measured Is What Matters: Targets and Gaming in the English Public Health Care System, in Public Administration 84(3), 2006, pp. 517–538. For an overview of the NHS target experience: The Health Foundation, On Targets: How Targets Can Be Most Effective in the English NHS, 2015.

³⁶ D. Ravitch, The Death and Life of the Great American School System: How Testing and Choice Are Undermining Education, Basic Books, 2010; L. Darling-Hammond, Evaluating ‘No Child Left Behind’, in The Nation, May 2007. For the Atlanta case: reporting in the Atlanta Journal-Constitution and the RICO convictions of 2015.

³⁷ J.A. Eterno and E.B. Silverman, The Crime Numbers Game: Management by Manipulation, CRC Press, 2012; J.A. Eterno, A. Verma and E.B. Silverman, Police Manipulations of Crime Reporting: Insiders’ Revelations, in Justice Quarterly 33(5), 2016. For the international diffusion of CompStat see the PERF reports.

³⁸ F. Zhang et al., GDP Manipulation, Political Incentives, and Earnings Management, in Journal of Accounting, Auditing and Finance, 2022; further World Bank, Bureaucrats, Tournament Competition, and Performance Manipulation: Evidence from Chinese Cities, Policy Research Working Paper 9938, 2022. For the broader frame of the Chinese promotion tournaments, see Chenggang Xu, The Fundamental Institutions of China’s Reforms and Development, Journal of Economic Literature 49(4), 2011.

³⁹ A. Nove, The Soviet Economic System, Allen and Unwin, 1977; J. Berliner, Soviet Industry from Stalin to Gorbachev, Edward Elgar, 1988. The nail and glass anecdotes originate in Krokodil cartoons of 1957 and 1960; Nove and Berliner provide empirical underpinning for the structural variants of the problem.

⁴⁰ W.N. Espeland and M. Sauder, Engines of Anxiety, op. cit. (note 7).

⁴¹ J. Huibers, The Sincere Voice, Statecraft Series Series III, Nº 01, 2026. For the elaboration of the Aiki method in the book corpus see De Richting van de Beweging, manuscript, chapter 11, with an explicit ethical compass: moving with the energy of a system works only when the intention serves the collective interest. Without that footing, the method loses its bearing and is put to work in service of the very pattern it ought to correct.

⁴² J. Huibers, De Richting van de Beweging, manuscript in preparation, chapter 9, Embedding as the First KPI.

⁴³ T.L. Saaty, The Analytic Hierarchy Process, McGraw-Hill, 1980; further the TOPSIS and ELECTRE traditions. For application in public policy design: V. Belton and T.J. Stewart, Multiple Criteria Decision Analysis: An Integrated Approach, Kluwer, 2002.

⁴⁴ O. O’Neill, A Question of Trust: The BBC Reith Lectures 2002, Cambridge University Press, 2002. The cited fragment is from the fifth lecture, Licence to Deceive.

⁴⁵ For an overview of the Buurtzorg model: M. Monsen and J. de Blok, Buurtzorg: Nederlands Pleidooi voor Vertrouwen, Empathie en Vakmanschap in de Zorg, self-published, 2013; and, internationally, S. Gray et al., Buurtzorg Nederland: A Global Model of Social Innovation, Change, and Whole-Systems Healing, in Global Advances in Health and Medicine 4(1), 2015. For current Dutch and international scale: Buurtzorg annual reports 2024 and 2025.

⁴⁶ Municipality of Veendam, Jeugd Expertise Punt: een andere weg in de jeugdhulp, municipal accountability documents 2023–2024; EenVandaag, In 10 jaar ruim 7.000 jeugdzorgaanbieders erbij, Veendam besloot in te grijpen, op. cit. (note 23). An independent evaluation of the Veendam arrangement is, at the time of writing, not available.

⁴⁷ NIVEL, annual research publications on administrative load in general practice and in specialist medical care; LHV and NHG, periodic reports on workload and registration burden.

⁴⁸ J. Hardij et al., Inkoopformule ‘open house’, op. cit.; BNNVARA Kassa, op. cit.; Pointer (KRO-NCRV) and Investico, series of publications 2020–2025 on implementation practice in youth care, social security and housing.

⁴⁹ J. Huibers, The Sincere Voice, op. cit.

⁵⁰ J. Huibers, Dissociated Organisations, Statecraft paper 2025, available via statecraft.nl.

⁵¹ J. Huibers, The Discriminating Eye, House of Viridian / Nourishment, April 2026, available at nourishment.houseofviridian.org.

Picnic Netherlands market-share figure 2024: Circana (annual overview of supermarket spending 2024). Group revenue and net loss: Picnic Group annual accounts 2024, filed with the Estonian commercial register (the holding is based in Tallinn); for the Dutch sub-result see the Dutch statutory annual accounts of Picnic Technologies B.V. and affiliated entities. Edeka stake: Picnic and Edeka press releases 2021 (expansion of shareholding to approximately 32 per cent), confirmed in subsequent annual accounts. ↩