Cross-border jurisdictional vacuum in social security
Document C — The visual architecture
Author: Eric Messa
[[PAGEBREAK]]
[[TOC]]
[[PAGEBREAK]]
Document C is the visual document for the System Audit project. It shows at a glance what is set out in various parts of the other documents: the structure of the network, the flow of information between organisations, and the mechanisms that cause failures.
Document C is not a repetition of the other documents in terms of presentation. It offers a distinct perspective that exists solely in diagrams. Whilst Document A tells the story, Document B organises the facts, and Document D draws the conclusions, Document C reveals the underlying architecture that underpins all those facts.
The document is structured as a series of chapters. Each chapter consists of a separate diagram accompanied by a short text. There is no continuous narrative, nor are there any lengthy analyses. Each diagram illustrates a single mechanism, a single pattern, or a single connection which can be found in the other documents, but never at a single glance.
Figure 0.1 — The five documents. Each document has its own function. Document 0 is the introduction and reading guide, Document A is the narrative audit, Document B is the data matrix, Document C is the visual architecture, and Document D is the synthesis. Document C illustrates the architecture that underlies everything.
Every chapter of Document C uses the same visual language. Once you are familiar with that language, you can read any diagram. Here is the vocabulary.
Belgium and the Netherlands are portrayed as two distinct worlds: rigid, closed systems that do not adapt their rules to accommodate a citizen who lives in both countries at the same time. Both worlds are composed of four concentric layers: implementation, oversight, politics and symbolism.
The family lives between two worlds. The family is not a sphere but a membrane: a flexible, transparent form that stretches and contracts according to the demands of the day. At the weekend, it is entirely in the Netherlands; on weekdays, it spans both worlds.
Figure 0.2 — The three channels. The nature of a channel determines whether data flows through, is held up or is completely blocked. Channel 1 is used for ex post checks and levies; it is not used to provide citizens with information in advance.
In the diagrams, connections are colour-coded according to what is happening:
Document C follows the human sequence. It begins with humanity, then moves on to Belgium (because the author is Belgian), then to the Netherlands, then back and forth between the two countries, then to Europe, then to those looking at Europe, and finally to the world, which is supposed to see the whole picture.
Figure 0.3 — The path through this document. Seven sections, in the order of human reality: from the body, through the two countries, to Europe, to the observer, and finally to the conclusion.
Document C is not intended to be read from start to finish in one go. Each chapter stands on its own. Readers can choose their own path.
For those who wish to understand the whole picture: start with Block I (Man), followed by the other blocks in order. In this way, the picture builds up step by step: from man, through the two countries, to Europe, to the observer, and finally to the conclusion.
If you are specifically interested in a particular topic, select the relevant chapter. Tied sales, the cascade, the UWV, CZ, the local authority, the Tax and Customs Administration, the European level — each has its own chapter.
If you want to see the full network, read Section I (Humanity) first, then the conclusion. Taken together, these two sections provide an overview of the structure at a glance.
Document C is situated between Document B and Document D. Whilst B sets out the facts and D draws the conclusions, C reveals the architecture that makes the failure possible. If you want the facts: B. If you want the conclusions: D. If you want to understand why the facts lead to these conclusions: C.
What this document illustrates is the essence of the power vacuum at a glance: one body, two worlds, three channels. And a citizen who has to hold their own in the gap between the two worlds.
This is Document C — The visual architecture. All organisations referred to in this document are legal entities. Natural persons are referred to by their role or their symbol. The facts are based on Document B. The conclusions are set out in Document D. This document does not constitute legal advice.
[[PAGEBREAK]]
Two rigid worlds. One membrane that stretches and contracts. Three channels, each with its own character. And within each world, four institutional rings that citizens must navigate in order to get anything done. This chapter sets out the playing field: who is where, how the layers are structured, and which actors operate within each layer.
In this dossier, Belgium and the Netherlands are referred to as two worlds: the World of Belgium and the World of the Netherlands. Each world constitutes a self-contained, sovereign legal and administrative system. For a citizen who spends their entire life within a single world, the internal links generally fit together seamlessly.
However, both worlds are rigid. They lack flexibility. They do not adapt their administrative boundaries when a citizen exercises their right to free movement. Each system assesses applications solely on the basis of national definitions, national medical criteria and territorial conditions — the ‘country of residence’ principle versus the ‘country of employment’ principle. The two systems physically adjoin one another at the national border and rub shoulders, but do not merge at all at the operational level.
Between these two rigid worlds lies the membrane: the family. Unlike the two national systems, the family has no fixed geometric shape. It is a transparent, irregular form that must stretch and contract to accommodate the incompatibility between the two systems. Whilst the institutions remain rigid, citizens are required to be completely flexible.
The diaphragm moves in a cyclical motion with two positions:
Inside the membrane there are three people, represented by symbols:
Each world is structured from the outside in, consisting of four concentric rings. Anyone who becomes stuck in the outermost layer and seeks protection further up moves from the outside in through these four layers:
Although the two worlds do not merge, there are three structural axes of connection between Belgium and the Netherlands. They reveal a fundamental asymmetry between what the state knows about its citizens and what the state provides for them.
Surrounding both these national spheres lies the EU’s outer layer. This layer comprises three clusters: official European and treaty-based institutions, cross-border mediation bodies, and the societal resonance layer (academia, strategic litigation, the media and UN human rights). The dotted line illustrates the fundamental problem with European social law: the EU proclaims the free movement of persons and coordinates this on paper (Regulation 883/2004), but has no dedicated enforcement body capable of breaking through a deadlocked national chain.
Figure I.1 — The game board. Two rigid worlds (Belgium, the Netherlands), each with four concentric layers (Implementation, Oversight, Politics, Symbolism). All actors are numbered — the legend below this figure gives the name and role for each number. In the centre, the membrane bridges the gap in two positions (extended on weekdays, retracted at weekends). The three channels illustrate the structural tension: thick green (automated), narrow orange (one-off), red dotted (manual).
For readers unfamiliar with Belgium and the Netherlands: listed below are all the organisations featured on the game board, organised by world and by layer. First, their public or private-law function, then their specific role in this case.
The actors on the game board are not isolated units. They are bound by fixed relationships of authority and data that define the space within which the family must operate. The crux of the power vacuum is that these relationships are structured in such a way that no single actor bears ultimate responsibility for the whole.
| Ratio | Formal authority | Actual practice | Impact on the family |
|---|---|---|---|
| Regulatory body ↔ implementing body BE NIHDI, CDZ → CM NL ZIN → CZ |
Regulatory authorities oversee the operators and may issue guidelines. | No mandatory mandate to intervene in individual cases. The case worker decides independently at the counter. | Complaints to the regulator do not resolve a specific obstacle. |
| Federal ministries ↔ federal ministries BE FPS SZ → FPS Finance NL SVB → Benefits |
Ancillary services with their own statutory mandates. | Each department makes its own decisions within its own legal framework. Certificates issued by one department are not binding on another. | Medical recognition does not automatically lead to tax protection. |
| Data flow: CM → KSZ → Parentia / SVB |
Statutory data exchange via the Cross-Sectoral Data Bank and the European EESSI network. | Parentia and the SVB are entirely dependent on the source code provided by CM. If CM fails to deliver, the entire chain comes to a standstill. | Family allowances and social benefits are being frozen without the recipient being able to do anything about it. |
| EU infrastructure Channel 1 FPS Finance ↔ Tax and Customs Administration NL |
Linked via EU directives (DAC) and the bilateral tax treaty. | Fully automated exchange of information on worldwide income. Maximum transparency for tax collection and recovery. | The EU knows all about income, but offers no safety net should it be lost. |
| Verdrag Canal Canal 2 CM ↔ CZ / CAK |
Linked via form S072 (Regulation 883/2004). | A narrow, one-off channel. It becomes problematic as soon as the care or status falls outside the standard form. | Ongoing friction over authorisations, parcel limits and personal contributions. |
| Medical sector Channel 3 Vitaz · UZ Gent ↔ ZorgSaam · Pallion |
No cross-border connection. | National EHR systems are hermetically sealed at the national border. | The family carries the files themselves. The patient becomes the sole link between two systems. |
| Politics ↔ implementation : Belgian Chamber of Representatives & Dutch House of Representatives |
Parliaments are the highest legislative bodies and exercise oversight over the government. | Politicians do not intervene in individual implementation cases. | Structural flaws in the law persist in its implementation. |
| EU ↔ Member States EU outer layer |
EU law takes precedence over national law (Regulation 883/2004). | The EU lacks its own enforcement body. SOLVIT is non-binding. ELA refuses to assist citizens. | A circular reference between Brussels and the national capitals. |
| Symbolism of the Royal Palace of Belgium and the Dutch Royal Family |
Constitutional head of state, with no executive powers. | Ceremonial postbox. Petitions are forwarded to the minister. | The end stop returns letters to their starting position in Layer 1. |
All organisations mentioned in this chapter are legal entities; their names are public and are listed in Document B. Natural persons are referred to solely by their role or by a symbol (♂, ♀, small ♂). This is both a methodological choice and a requirement under the General Data Protection Regulation. The author of this report is the patient in question; he is permitted to publish his own medical and administrative data.
[[PAGEBREAK]]
Block II will shortly show the pots, the mechanism and the decision. This chapter shows what happens when the mechanism and the decision converge on a single family, in a single timeline, with one visible consequence: an empty bank account.
11 April 2025 · Week 1
The body is admitted. The medical report and the invoice are separated: the report remains at the hospital, whilst the invoice is sent to the health insurance fund. The patient’s contribution is deducted directly from their bank account.
ongoing · from day 1
One line continues to flow uninterrupted: the mother. Everything the father contributes comes to a halt.
April 2025 · ongoing
Whilst the mother’s wages come in, they leak out at the other end. Small and predictable, or large and sudden. The reserve is running dry.
May 2025 — September 2026 · blocked
Five channels would need to be in place to offset the loss of income. Four are blocked. One small channel — child benefit — keeps the family just above the poverty line.
autumn 2026 · eighteen months later
Eighteen months later, the account is in the red. The commercial bank has offered a solution involving penalties. The debt collection agency has been called in. And the system that was supposed to have dealt with this is still not delivering any results.
The facts are based on Document B. Amounts can be found in the relevant entries for each organisation. Natural persons are identified by symbols (♂, ♀); organisations by their public name.
[[PAGEBREAK]]
A family that loses its income due to illness can apply for a range of benefits in Belgium and the Netherlands. These are spread across two countries, two tax systems, two social security systems and two benefit schemes. This chapter first sets out what funds are available. It then explains why most of these funds remain untapped — not due to a single cause, but because of a mechanism that is common to both countries.
In theory, the family is entitled to, or may claim, the following benefits. The status for each fund is based on Document B.
| Pot | Administrator | Condition | Status |
|---|---|---|---|
| Sickness benefit (incapacity for work, first year) | CM | Declared unfit for work + waiting period completed | Blocked — 180-day waiting period not applied under the European aggregation rule |
| Disability benefit (after 1 year of incapacity for work) | CM | Automatic transition after 365 days of incapacity for work | Blocked — fictitious end date prevents transfer |
| IVT (income replacement allowance) | FPS Social Security / DG HAN | Recognised disability + reduced earning capacity + residence in Belgium | Blocked — residence requirement, no Belgian domicile |
| Increased tax-free allowance | FPS Finance | Recognition of a severe disability (9 points) | Blocked via DG HAN, activated via a tax bypass (see Chapter II — The Belgian tax route) |
| Child Benefit / Growth Package | Parentia | Dependent child + country-of-employment principle (mother works in Belgium) | In work — main payer registered in the mother’s name, supplemented by the SVB |
| Social allowance (as part of the Groeipakket) | Parentia | Household income below the threshold + 6 consecutive months | Blocked — Parentia assesses on the basis of theoretical earning capacity, not actual income |
| MAF (maximum invoice amount) | CM | Co-payment above the ceiling — increased ceiling upon recognition of a disability | Partially — standard MAF active, enhanced MAF blocked due to disability refusal |
| Tax relief on medical expenses | FPS Finance | Proof of medical expenses via the annual tax return | Partially — tax-deductible, but with limited impact as long as other income remains low |
| Assistance with medical aids | CM (under the NIHDI agreement) | Medical requirements and approval by the consulting doctor | Works — orthopaedic shoes approved, diabetes convention in progress |
| Diabetic Foot Clinic, UZ Gent | CM (convention) | Agreement with a teaching hospital | Valid — approved from 19 June 2026 to 27 May 2027 |
| European parking permit | FPS Social Security | Recognition of disability | Granted — 3 August 2026, indefinite duration |
| Public transport discount card | FPS Social Security | Recognition of a disability + specific mobility barrier | Rejected — 21 August 2026, medical criteria not met |
| Certificate of Fitness to Drive | FPS Mobility / ophthalmologist | Visual acuity above the statutory threshold | Rejected — visual acuity below the threshold, permanent |
| Jar | Administrator | Condition | Status |
|---|---|---|---|
| WIA benefit (incapacity for work) | UWV | 104-week qualifying period + insurance cover on the reference date | Blocked — incorrect reference date, Vester judgement not applied |
| Social assistance benefit (Participation Act) | Municipality of Hulst | Insufficient income + assets below the threshold | Blocked — request for help deleted after 73 days without being read |
| Special assistance | Municipality of Hulst | Special expenses + income below the threshold | Blocked — no substantive assessment |
| Healthcare allowance | Benefits Department | Insured in the Netherlands + income below the threshold | Blocked — The RIV register does not contain any treaty policies |
| Housing benefit | Benefits Department | Rent above the threshold + income below the threshold | Partially — granted, amount revised on the basis of incomplete income details |
| Child-based budget | Benefits Department | Child + income below the threshold | Partially — granted following SVB linkage |
| Child benefit | SVB | Dependent child + country of residence principle | Granted — small amount, active since 20 April 2026 |
| Claim | Basis | Condition | Status |
|---|---|---|---|
| Aggregation of insurance periods | Regulation 883/2004, Article 6 | Periods of insurance in several EU Member States | Not applied — CM and UWV apply national reference periods |
| Pro-rata calculation | Regulation 883/2004, Article 52 | Entitlement to benefits under the aggregation rule | Not applied — condition: recognition of entitlement to benefits |
| Application of the Vester judgement | ECJ C-134/18 | Income gap arising from successive insurance policies in two Member States | Not applied — UWV refuses, CM refuses |
Figure II.1 — The inventory. Surrounding the family are the benefits to which they are entitled or might be entitled. Green = active. Orange = partially active or via a roundabout route. Red = blocked or not claimed. On the left are the Belgian funds, on the right the Dutch ones. The green funds are small or indirect. The large income funds are all red.
The schemes are not independent of one another. In both countries, there is a link between them: a medical assessment leads to two possible outcomes — a benefit payment and a tax relief — but the tax relief is only granted once the entitlement to the benefit has been recognised. Legally, they are two separate processes. Administratively, they are linked.
Figure II.2 — Tied sales. One medical assessment, two routes. The tax route is not assessed independently, but is linked to the benefit route. Applied in both countries, with their own administrative variations.
A national citizen has a single system. If they encounter problems with tied sales, they can take legal action through a single point of contact. A cross-border worker has two national systems which do not recognise them as a single person.
In your country of residence (the Netherlands): you live here, so we are not authorised to pay the benefit, as you were working in Belgium. In your country of employment (Belgium): you work here, but you live in the Netherlands, so we do not have to grant the tax benefits. Each system applies its own tie-in, and no system resolves this by decoupling it from the medical assessment.
The EESSI network and Regulation 883/2004 provide for the aggregation of periods (Article 6) and pro rata calculation (Article 52). However, they do not provide for the decoupling of the medical assessment from the conditions for entitlement to benefits. Such decoupling falls within national competence, and neither state applies it to cross-border workers.
In practice, tying leads to a series of circular arguments. Below are the four most common ones.
CM has refused because the national waiting period has not been met. The UWV is awaiting a decision from the Belgian authorities before applying the Vester judgement. The citizen is caught between two authorities, each of which imposes its own condition that the other authority has made impossible to fulfil.
One federal body grants formal recognition of a disability. Another federal body imposes a stricter threshold for tax purposes. The citizen must reopen the case with the first body in order to score one more point — for a physical condition that has already been established by five independent healthcare professionals.
The CM does not record sickness status in the Cross-Sectoral Social Security Database. Parentia, the SVB and the Dutch Benefits Agency all access data via this same database or a derivative thereof. A single block on data from one agency halts six other benefit payments.
The health insurance fund administers both the maximum invoice (MAF) and the recognition of disability. Anyone recognised as disabled is subject to a lower excess threshold — lower co-payments, higher reimbursements. Because the same health insurance fund refuses to grant this recognition, the family remains subject to the standard threshold. The patient therefore pays more co-payments than a person with a disability in the same physical condition. The same body has established the medical reality through the approval of medical aids, but refuses to use that reality to increase the financial safety net.
The family is entitled to a wide range of benefits. In theory, the combined network of the two welfare states covers almost every aspect of their medical and financial situation. In practice, however, more than half of the funds remain inaccessible. Not because the eligibility criteria are impossible to meet, but because the pathways to accessing them are structured in such a way that they block one another.
The linking of the benefits route and the tax route is the most common mechanism. The catch-22s are the result of this: circular arguments between two national systems that do not communicate with one another, and between federal authorities within the same system that do not recognise each other’s thresholds.
The following chapter focuses on the actors that perpetuate these catch-22s — not as individuals, but as roles within a system.
All organisations in this chapter are legal entities. Natural persons are referred to solely by their role or by a symbol (♂, ♀). This is both a methodological choice and a requirement under the General Data Protection Regulation. The ‘pots’ inventory is based on Document B, which records the actual applications, statuses and refusals for each organisation.
[[PAGEBREAK]]
Five healthcare professionals examined the body. Two medical officers made the decision. This chapter shows that both bodies — the Belgian health insurance fund and the Belgian Federal Public Service — apply the same mechanism, each with its own variation.
Five healthcare professionals, working in two countries, examined the body. All five identified the same permanent disability.
Figure III.1 — Two variants, one mechanism. The health insurance fund (variant A) allows its own doctor to make the decision without examining the patient. The federal public service (variant B) allows its own doctor to make the decision on the basis of a measurement that falls one point below the tax threshold. In both cases, the administrative threshold takes precedence over the medical assessment.
In both cases, the doctor making the decision is on the agency’s payroll. In both cases, an administrative threshold is applied which does not exist in clinical practice. In both cases, five independent medical assessments carry less weight than a single administrative decision.
That is not the personal fault of any individual doctor. It is the system: the doctor carrying out the assessment is not a better doctor, has not undergone more training, nor does he possess deeper knowledge. He is the gatekeeper of an administrative hurdle. The gate is closed, not because the body isn’t unwell — that has been proven five times over — but because the algorithm behind the gate says: the benefit isn’t being paid, so the recognition isn’t granted, so the tax route isn’t available.
All organisations mentioned in this chapter are legal entities. Natural persons are referred to solely by their position. The facts are based on Document B.
[[PAGEBREAK]]
An organisation that accepts the medical reality, but then stumbles over its own revenue model. Three levels, one outcome. And the pattern repeats itself at another organisation.
September 2025 — April 2026
April — September 2026
September 2026
The medical route has been cleared. The evidence has been accepted. It then transpires that the income model does not reflect the actual situation. The figures add up, but the family falls outside the criteria.
Municipality of Hulst · 2026
What happened at Parentia is happening again at the local authority. Each organisation has its own medical portal and its own second hurdle. Citizens have to go through the same process all over again.
Facts based on Document B. Natural persons are identified by symbols; organisations by their public names.
[[PAGEBREAK]]
So far, this case study has mainly highlighted what goes wrong. This chapter shows the opposite: the Belgian tax route works, and provides the family with a substantial annual benefit. Four tax deductions and allowances, one of which was achieved via a workaround. This workaround has now been confirmed twice — once by the submission of the tax return itself, and once by an independent written confirmation from the tax authorities that the chosen approach is legally sound.
| Tax benefit | Basis | Status in this case |
|---|---|---|
| Increased tax-free allowance (disability code) | Sections 131 and 135 of the WIB 92 · permanent disability of ≥66% · earning capacity less than one-third | Activated via a bypass, confirmed twice. Benefit: €2,669.97 on the joint tax return for 2026. |
| Marriage ratio | Section 87 of the WIB 92 · allocation of part of the earned income to the partner with the lower income | Applied automatically when filing a joint tax return. Reduces the tax burden on family income. |
| Childcare tax relief | Art. 14535 WIB 92 · costs for the care of children under the age of 14 | The holiday childcare certificate has been added manually to the 2026 tax return. |
| Tax relief on medical expenses | Sections 140–145 of the WIB 92 · medical expenses exceeding a certain percentage of income | Partially tax-deductible. Limited impact as long as income remains low. |
Figure VI.1 — The four tax relief streams. Three green streams: increased tax-free allowance, marriage allowance and childcare allowance. One orange stream: medical expenses (threshold). Together, they determine the tax relief shown on the annual Belgian tax return.
Three roadblocks on the main road:
Figure VI.2 — The tax bypass. Three obstacles on the standard route. The tax authorities are legally separate from the benefits system and are permitted to accept their own medical evidence. A single document from the GP, drawn up in the language of the tax authorities, triggers the increased tax-free allowance without the need to amend the central database.
The bypass is based on a single assumption: that the tax authorities will accept a medical certificate from the GP without the involvement of the FPS Social Security. This assumption has been confirmed in this case in two independent ways.
On 25 September 2026, the final Non-Resident Tax Return (tax year 2026, income year 2025) was submitted by the ACV’s tax department. The return includes:
The result: €2,669.97 in the family’s favour. The tax return was submitted on 25 September 2026 with a calculated balance; we are awaiting final processing and the tax assessment notice from the FPS Finance.
In parallel with the submission, an independent researcher specialising in cross-border employment — employed at an academic institution and not involved in the case — made an anonymous enquiry to the FPS Finance. The question was whether the voluntary inclusion of medical certificates with the tax return is a valid means of demonstrating tax disability status under Belgian law.
The FPS Finance, Non-Resident Tax Management and Services Team (PG15 Ledeberg), has confirmed this in writing. In two successive replies, the tax authorities state that:
The result is that not only does the bypass work in practice, but the tax authorities themselves confirm that the chosen approach is legally sound. It is no longer a matter of guesswork or favour. It is formal confirmation that the approach the family has found is the right one.
Figure VI.3 — The double confirmation. The same bypass route has been confirmed twice: once in practice (the final tax return submitted, showing the calculated benefit), and once in theory (the tax authorities’ written confirmation to an independent investigator).
The Agency for Educational Services (AGODI) is the Flemish agency responsible for the payroll administration of teaching staff. In this case, AGODI acts as the mother’s employer. When the federal tax benefits are activated, AGODI must reflect them in the payroll records. And that is where things initially go wrong.
AGODI rejects the retroactive amendment to the payroll tax regarding the teachers’ exemption under the double taxation agreement. The reason: the tax statements for previous years have already been issued.
AGODI refuses to make the adjustment on the basis of the federal certificate awarding 8 points. The payroll department has taken it upon itself to carry out a medical and tax assessment: the certificate is said to be valid only for a parking permit, and 9 points are normally required for a reduction in payroll tax.
Once the conclusive medical certificate (permanent disability ≥66 per cent, earning capacity reduced to one third) has been submitted, AGODI takes over completely. The authority suddenly states that it does not need to receive any supporting documents, that the school must send only an electronic adjustment, and that the tax authorities will ultimately settle the account.
On 23 September 2026, the school secretariat recorded the official notification ‘change in tax status of disabled partner’ in the Informat staff system (notification number 4, [partner’s registration number]). This electronic notification records, with retroactive effect from 1 January 2026, that the dependent partner has no earned income and is formally registered as disabled. The administrative process has thus been fully completed.
Figure VI.4 — From a moving target to final completion. Three rounds of rejection, followed by the final registration of the electronic signal. What initially seemed impossible has now been administratively finalised.
The Belgian tax route works — perfectly. Four tax deductions, one bypass, a joint tax return offering a substantial benefit. The workaround has been confirmed twice: once in practice, and once by the tax authorities themselves. And the final authority in the chain (AGODI) ultimately issued the electronic signal that registered the tax status with retroactive effect.
What this chapter shows is that, within the very same system that structurally excludes cross-border workers in other areas, there are indeed avenues that do work. The tax route is one such example. It does not depend on health insurance registration, nor on the disability services, nor on European coordination. It runs directly from the GP to the tax authorities.
And just as importantly: the route has been confirmed by the tax authorities themselves and verified by an independent investigator. The outcome is not the result of any favour. It is the result of the correct application of the law by a member of the public who has found a way to circumvent the blocked channels.
The amounts are based on the 2026 tax return and may vary from case to case. The relevant statutory provisions are Articles 87, 131, 135, 140–145 and 14535 of the Income Tax Code 1992. All organisations mentioned in this chapter are legal entities. Natural persons are referred to by their position or their symbol. This chapter does not constitute tax advice.
[[PAGEBREAK]]
Anyone who receives a decision from the Federal Disability Service and disagrees with it will find in the letter that they can take the matter to the employment tribunal. What the letter does not mention is that there are four other options – cheaper and quicker – which the individual can pursue themselves.
The Federal Disability Service’s website lists four options that members of the public can pursue themselves — without a solicitor and without going to court. These are known as ‘re-examinations’. Applications can be made in writing via the contact form or by post.
Figure VII.1 — The four routes to reopening the case. These are set out on the website of the Federal Disability Service, not in the letter. Applications for them can be made in writing, without a solicitor and without going to court.
A request for a review is not an appeal to the court. It is a request to the same body to reconsider the decision — with new arguments, new evidence, or an explanation of why the previous decision is incorrect. The attending doctor can assist with this: he or she can explain in writing why the disability service’s decision is incorrect.
Unlike a court case, reopening a case does not require a solicitor or court fees. It simply involves a letter or an online form. And the decision is again made by the same body, but this time taking into account the arguments put forward by the treating doctor.
The information in this chapter is based on the official page ‘Applying for a reopening’ of the FPS Social Security (DG People with Disabilities), last updated on 15 September 2025. The legal basis and the procedure can be found at handicap.belgium.be. This chapter does not constitute legal advice; if in doubt, consult a specialist service or a solicitor.
[[PAGEBREAK]]
The health insurance fund is not a bank with a single account. It is an organisation with three separate streams of funding: compulsory health insurance, the CM package of services and benefits, and optional insurance schemes. The first stream comes from the government via the RSZ. The second from membership fees. The third from separate premiums. And each of these streams feeds a different department. This chapter explains where the money comes from, why one branch works and another does not, and who can ultimately intervene when the CM refuses.
Figure XIV.1 — The three funding streams. Compulsory health insurance is funded by social security contributions and taxes, administered by the NIHDI, and delivered by the health insurance funds. CM services and benefits are funded by membership fees. Voluntary insurance schemes are funded by separate premiums. Each of the three funding streams feeds a different service.
Compulsory health insurance is not paid directly by the CM. The employer deducts a portion of the wages and pays it to the National Social Security Office (RSZ). The RSZ distributes the money amongst the various social security branches. There are two separate branches for health insurance:
The NIHDI receives the total budget and allocates it amongst the health insurance funds. Each health insurance fund receives a budget based on the number of members and their profile. The health insurance fund uses this budget to pay for services provided to its members. The health insurance fund receives a separate reimbursement to cover administrative costs.
Figure XIV.2 — The path from wages to healthcare. The employer deducts the RSZ contribution from wages. The RSZ pays the funds to the NIHDI. The NIHDI distributes the funds amongst the health insurance funds. The health insurance fund makes payments from two categories: medical care and benefits.
Figure XIV.3 — Two branches, two sets of criteria. The healthcare branch operates on the basis of objective, measurable performance. The benefits branch operates on the basis of subjective assessments and national criteria. The same institution, the same National Social Security Fund (RSZ) contribution, but a fundamentally different logic — and therefore a fundamentally different outcome.
The Health Insurance Funds Supervisory Authority (CDZ) oversees the CM as a whole — including compulsory insurance, supplementary services and optional insurance schemes. The CDZ audits the accounts, monitors compliance with financial rules and ensures the correct application of statutory provisions. It may issue recommendations to the Minister and approve amendments to the articles of association.
What the CDZ does not do: settle individual disputes. The CDZ is an administrative supervisory body, not a dispute resolution body. It can investigate the structural functioning of the health insurance fund, but cannot review a decision in an individual case. It is not an appeals body.
When an implementing body systematically disregards the law, there is a judicial body in Belgium which, in theory, can intervene: the Labour Audit Office, the public prosecutor’s office attached to the labour courts. It is responsible for enforcing social security legislation and can take action when an organisation fails to comply with the law.
On 10 September 2026, a formal report was submitted to the Ghent Labour Inspectorate in this case. The report concerned systemic procedural errors at the CM, a breach of the European Coordination Regulation, and a failure to comply with NIHDI guidelines. The full timeline and the medical certificate for tax purposes were attached as supporting documents.
Figure XIV.4 — The Labour Audit Office. The report reaches the Labour Audit Office on 10 September 2026. The response comes one day later: the body declares that it has no jurisdiction and classifies the structural deadlock as a civil dispute. The citizen is referred to the employment tribunal, the slowest route.
What the Labour Inspectorate is doing here is the same as what the CDZ does: passing the buck on the issue of responsibility. The body that is supposed to intervene in the event of a breach of social legislation classifies the dispute as a private matter. And the public is left with one less body to turn to.
The Medical Council is the statutory body responsible for upholding medical ethics and exercising disciplinary authority over individual doctors. It can investigate complaints and impose disciplinary sanctions — ranging from a warning to striking off the register of doctors — but has no authority over administrative decisions made by institutions or over algorithmic processes within a health insurance fund.
In this case, a complaint was lodged with the Provincial Council of Antwerp on 19 September 2026 regarding the conduct of the CM’s medical adviser. The complaint concerns the issuing of conflicting decisions bearing a visually identical signature, without traceable access to the current medical file. The question is whether this practice is ethically compatible with the independent practice of medicine.
Figure XIV.5 — The Medical Council. Disciplinary proceedings may affect the individual doctor, but not the framework that produces the decision. Even a well-founded complaint does not alter the outcome for the patient, as the health insurance fund replaces the doctor and the algorithm remains unchanged.
The question of why medical aids are covered but benefits are not has a structural explanation. The CM carries out two legally distinct tasks within the compulsory health insurance scheme. One task is based on measurable outcomes and fixed rates. The other is based on subjective assessments and national criteria.
There are therefore two different systems within the same organisation. When it comes to medical care, the CM does not need to assess an individual — it simply pays the bill. For benefits, however, it must assess whether someone is unfit for work, and that assessment is subject to conditions laid down by law. For a cross-border worker, these conditions are problematic, as they are based on national reference periods and residence requirements.
And when the CM refuses, the two supervisory bodies — the CDZ and the Labour Inspectorate — do not consider themselves competent to deal with individual cases. The Medical Council can only take action against the individual doctor, not the system that drives the decision. This conclusion is not an accusation of bad faith. It is simply an observation that the system does not assign responsibility for the whole to any single body. Each body carries out its own part. And no single body bears responsibility for the whole.
All organisations in this chapter are legal entities. Natural persons are identified by their role or their symbol (♂, ♀). The information on funding is based on official communications from the CM, the NIHDI and the CDZ. The procedural information concerning the Labour Inspectorate and the Medical Council is based on Document B and public sources. This chapter does not constitute legal or financial advice.
[[PAGEBREAK]]
The UWV assesses whether a person is unfit for work and whether they are entitled to benefits. What follows is not a single refusal. It is a series of refusals. Each refusal has its own reason, its own procedure and its own outcome. None of the refusals is examined in substance. And NEO, the Belgian equivalent, in turn employs the same tactic in the opposite direction.
Figure IX.1 — The six rounds. Five applications, two objections, one review. Each round ends with a formal rejection. At no point throughout the entire process is the medical situation assessed in substance, despite five independent medical assessments and hundreds of pages of evidence.
The UWV is not the only body that blocks the issuance of a career certificate. On the Belgian side, the National Employment Office (NEO) employs a similar tactic, but in relation to teaching careers.
The partner works as a permanent teacher in the Flemish education system. To apply for the European Career Certificate (PD U1), the NEO must complete a form documenting their entire teaching career. However, the NEO is blocking the application by requiring a C4 form.
What is the C4 form? It is a termination document used in the private sector. It is used when an employee in the private sector leaves their job. For permanent staff in the education sector, this form does not exist in legal terms. There is no termination, no private employer, no C4.
NEO is therefore applying the logic of the private sector to a career that falls entirely outside that framework. The agency is forcing a public-sector career into a private-sector mould.
NEO requires a C4 form for a career path for which there is no legal basis for a C4. The application is blocked. The citizen must provide alternative evidence themselves via the Agency for Educational Services (AGODI), which issues the official Calculation of Financial Seniority.
The patient resolves the issue by submitting the AGODI certificate in person. The Agency for Educational Services issues an official document setting out the patient’s full educational history. This document is sent to NEO, after which the second PD U1 certificate is issued.
This results in a third layer of manual intervention. The patient acts not only as a data hub between two countries, but also as a data hub between two agencies within the same country: between the federal NEO and the Flemish AGODI. These two levels of government, which have access to the same career history, do not automatically exchange this data.
Following the issue of the second PD U1 certificate, the patient submits a formal request to synchronise their Dutch employment records with the Belgian MyCareer system. NEO refuses.
The UWV provides the data, but the NEO refuses to process it. The patient bypasses the NEO via another federal agency. Two government agencies within the same country cannot exchange data with one another without the citizen acting as a data hub.
What the UWV consistently does is not to refuse. It is to reject on procedural grounds. That is a fundamental difference.
In the event of a refusal on the merits, the UWV would have to state that the medical condition is not severe enough. This would mean that a doctor had examined the patient, confirmed the amputations, tested their visual acuity, and, on that basis, concluded that the person was not entitled to benefits.
In the event of a procedural rejection, the UWV states: the application does not meet the formal requirements. No tick. Incorrect reference date. No basis for insurance. No Belgian decision. No C4 form.
A procedural dismissal has three advantages for the authority:
The UWV and the NEO follow the same pattern here, but in the opposite direction. The UWV refuses to take note of the Belgian decision. The NEO refuses to take note of the Dutch decision. Both bodies are waiting for the other, and neither is taking responsibility.
Figure IX.4 — Six rejections, six reasons. All the reasons are procedural. None of the reasons are medical. The medical situation is not assessed in substance at any stage of the process.
4 June 2026
On 1 June 2026, the patient submits three additional supporting documents for the first appeal: the Caren report from the Dutch home care service (126 clinical entries confirming the period of hospitalisation from 11 April to 9 May 2025), an ophthalmological report from the teaching hospital, and an endocrinological report.
Three days later, the UWV declared the appeal unfounded. This was six weeks before the deadline of 15 July 2026, as announced by the UWV itself. The Caren report — the Dutch evidence confirming the reference date — was not examined in substance.
8 July 2026
The second objection concerns the WIA rejection of 14 April 2026. The patient cites the Vester judgement: in the case of successive insurance schemes in two Member States, a Member State may not apply the qualifying period in such a way as to create an income gap. The UWV explicitly rejects this, arguing that the loss of income cannot be determined without a formal decision from Belgium.
As Belgium is also refusing, the UWV closes the case with the message that the citizen must first secure a decision from the Belgian authorities. The responsibility for breaking the deadlock is shifted back onto the citizen.
The UWV and NEO follow the same pattern, but in the opposite direction. Both agencies reject claims on procedural grounds. Both agencies refuse to take the other’s information into account. Both agencies pass the responsibility back to the public.
The member of the public is caught between two bodies, each of which imposes its own condition that the other body has made impossible to fulfil. And throughout the entire process, the medical situation is not assessed in substance even once.
All organisations mentioned in this chapter are legal entities. Natural persons are referred to by their role or their symbol (♂, ♀). The facts are based on Document B. This chapter does not constitute legal advice.
[[PAGEBREAK]]
CZ administers the treaty-based insurance policy for a resident of the Netherlands who is employed in Belgium. CZ reimburses the cost of healthcare, but refuses to register the person for the healthcare allowance. CZ also refuses to provide patient transport, citing a requirement that does not exist in the country of treatment. In both cases, the organisation’s protocol takes precedence over medical reality. And in both cases, the authorities that should be ensuring the process runs smoothly are failing to do so.
A treaty policy is a European instrument. The employee lives in the Netherlands and works in Belgium. Under the European coordination rules, the country of employment is responsible for social security. The employee pays their social security contributions in Belgium, to the Belgian health insurance fund.
However, he is also entitled to healthcare in his country of residence. That is what the treaty policy is for: CZ acts as the Dutch administrator of the policy, reimburses the Dutch healthcare costs, and settles them with Belgium.
The premium goes to the country of employment (Belgium, via CM). CZ reimburses Dutch healthcare costs but does not receive a premium. Registration in the RIV register involves a lot of work for CZ without any corresponding income.
The result is that, in the Netherlands, the holder of a treaty policy is insured for healthcare, but not for the benefit. The Benefits Service refers exclusively to the RIV register. If it isn’t in there, it doesn’t exist.
CZ’s refusal to register the treaty policy is not the only issue. The RIV register — the Reference Database for Insured Persons under the Health Insurance Act — is managed by the Central Administration Office (CAK). And the Dutch Healthcare Institute (ZIN) oversees its implementation. In this case, all three bodies involved are refusing to resolve the problem.
Figure X.2 — The closed triangle. CZ refuses to register the data. The CAK manages the register but refuses to synchronise it. The ZIN supervises the process but refuses to grant access to individual files. Each body points the finger at the other. The citizen is trapped in a vicious circle that nobody is breaking.
The CAK manages the Health Insurance Act Insured Persons Reference Database (RIV) and administers the contractual schemes. If, in April 2026, the patient formally demands that the active and paid-up treaty policy be entered into the register, the CAK will expose a fundamental architectural flaw in its own systems.
The authority refuses to register the policy on procedural grounds. The CAK states that it only registers ‘ordinary’ cross-border workers: those who work in the Netherlands and live abroad. For the reverse situation — living in the Netherlands and working abroad — there is simply no category programmed into the system.
The system does not recognise the form field. And rather than adding the field, the CAK closes the file.
The Dutch Healthcare Institute (Zorginstituut Nederland) manages and supervises the RIV register. When, in April 2026, the patient asks the Institute to investigate and rectify the exclusion from the register, the Institute states that it has no access to individual personal data. It refuses to check the register and refers the matter back to CZ and the CAK.
The body that manages the register cannot access the register. That is the essence of the closed triangle: each body has a limited scope of authority, and no single body has the full scope.
Due to amputations and loss of vision, the patient is no longer able to drive independently. His healthcare providers are in Belgium. He lives in the Netherlands. He requires adapted transport for these journeys — seated patient transport. In the Netherlands, this is a benefit provided through the health insurer. For a policy under a treaty, this is CZ.
On 18 March 2026, the patient submits an application for authorisation for seated patient transport to the Belgian hospital. On 25 March, CZ sends a request for further information: a perimetry test — an examination of the field of vision — is required.
What is perimetry? An ophthalmological examination in which the full perimeter of the visual field is measured. It is a standard examination in Dutch ophthalmology. It is not a standard examination in the Belgian protocol for this specific condition — diabetic retinopathy with macular oedema. The Belgian hospital does not carry out this test on this patient, as it is not included in the treatment protocol.
The result is that the patient cannot provide the evidence, as the test is not available in the country where he is being treated.
CZ is requesting a test that is not routinely carried out in the country where the patient is receiving treatment. The patient is unable to provide the evidence, not because he is unwilling to do so, but because the test is not included in his clinician’s treatment protocol.
Figure X.4 — The dance. Five rounds, three rejections, one escalation. A new procedural requirement each time. Granted each time. And each time, the threshold shifts. The patient provides the evidence, but the protocol does not accept it.
Figure X.5 — The parallel. CM and CZ apply the same logic: medical evidence is accepted where it fits within their own protocol, and rejected where the protocol imposes a procedural requirement that the patient cannot meet. The protocol is the truth, not the body.
In both cases, the patient has provided the maximum evidence he is able to provide. In the first case: a valid insurance policy, paid premiums, and an administering body in Belgium. In the second case: a complete medical file detailing amputations, visual acuity measurements, and an academic report from a university hospital.
In both cases, CZ rejects the claim on procedural grounds. The treaty policy is not a Dutch insurance policy. The scope is not specified. The mileage calculation is incorrect. The old letter carries more weight than the new report.
The crux of the matter is not that CZ is making mistakes. The crux of the matter is that the CZ protocol takes precedence over medical reality. And because the law allows CZ the leeway to follow the protocol, there is no breach. Just an outcome that the family cannot bear. And the CZ-CAK-ZIN triangle closes in on the family without a single loophole.
All organisations mentioned in this chapter are legal entities. Natural persons are referred to by their title or the symbol (♂). The facts are based on Document B. This chapter does not constitute legal advice.
[[PAGEBREAK]]
In the Netherlands, the local authority is the first point of contact for anyone facing financial hardship. For a cross-border worker in Zeeuws-Vlaanderen, the municipality of Hulst would be the logical place to apply for social assistance, special social assistance or social support. In practice, however, things are different. The local authority receives the cry for help, but does not provide any money. It offers equipment and assistance. And it uses calculation models that fail to reflect the actual situation. A formal appeal is now under consideration — but the decision on it has been postponed yet again.
Under the Dutch social security system, the local authority has three tools at its disposal to address income hardship:
These three schemes are designed to provide support when income is lost or insufficient. For a cross-border worker facing a medical emergency and a complete loss of income, that is precisely the situation for which they were created.
Figure XI.1 — The first helpline that doesn’t help. The local authority received six formal reports, spread over eleven months. Each time, the response is either a procedural one or no response at all. Applications for social assistance and special assistance are not assessed. The only formal procedure currently underway is the appeal lodged on 22 July 2026 — and the decision on that has been adjourned.
The only formal procedure remaining is the appeal lodged on 22 July 2026 against the decision of 22 June 2026 concerning special assistance. This is the first point at which the local authority is required to formally assess the application. The local authority responded by extending the decision-making period.
Figure XI.2 — The pending appeal. The appeal against the decision on special assistance was lodged on 22 July 2026. The local authority confirmed receipt on 21 September 2026 and extended the decision-making period by six weeks. The applicant did not request an oral hearing, as the grounds had already been sufficiently set out.
The reason why the local authority, like Parentia, does not allocate any funds is down to the calculation model. Social security benefits and special social security benefits are means-tested against an income threshold. In principle, that makes sense. But the model does not take the actual situation into account.
The calculation model uses historical tax data. It does not take into account that actual income has fallen to zero, that medical costs have risen, and that the household is living below the poverty line. It is exactly the same pattern as with Parentia.
The local authority does offer something: mobility. There are two schemes: the disabled parking permit (GPK) and the Wmo pass for adapted transport. Both work — to some extent. And both come with their own personal contribution.
The patient holds a European parking card issued by the Belgian federal government. The municipality of Hulst refuses to accept it and is demanding a new local assessment, with the associated administrative fees. What is striking is that the patient was required to resubmit the medical file on 17 September 2026, even though it had already been submitted on 7 March 2026 and 20 August 2026.
The Wmo card enables patients to get about. However, the personal contribution falls due during a period when they have no income. And anyone who fails to pay on time loses their card.
The system allows journeys to academic hospitals across the border, but refuses journeys to Belgian government agencies at exactly the same distance. Requests for assistance in cases of vision loss are also not recorded.
Figure XI.5 — Three bodies, one pattern. Parentia, the municipality of Hulst and CZ all apply the same mechanism: medical reality is recognised where it fits within the protocol, and rejected where the protocol imposes a procedural requirement that the citizen cannot fulfil.
In theory, the local authority is the first point of contact. It knows its residents, is familiar with the local situation and has the tools to intervene quickly. In this case, however, the local authority failed to do so. It received the distress call six times and issued six procedural responses. Not once was the financial situation assessed in any substantive way.
And the crux of the matter is this: in theory, the benefits should be enough. The Belgian sickness benefit, the Dutch WIA, the top-ups, the family allowance — together, they should be enough to keep a household afloat. The problem isn’t that the money isn’t there. The problem is that nobody is plugging the gap that has arisen between the various schemes. Medical costs are the trigger, but the administrative deadlock is the cause of the impoverishment.
The only formal procedure currently underway is the notice of objection dated 22 July 2026. The deadline for a decision on this has been extended by six weeks. The applicant has not requested an oral hearing, as the grounds have already been sufficiently set out. The decision will follow at a later date. Until then, the financial situation remains unchanged.
All organisations mentioned in this chapter are legal entities. Natural persons are referred to by their title or the symbol (♂). The facts are based on Document B. This chapter does not constitute legal advice.
[[PAGEBREAK]]
The Benefits Service pays out three means-tested benefits: healthcare allowance, housing benefit and child-related budget. In this family, one person is on full benefits, one is on partial benefits, and one is in work. And the tax return itself presents the taxpayer with a task they cannot carry out: entering a ‘worldwide income’ that they do not know, whilst being held personally responsible for its accuracy.
Figure XIII.1 — Three allowances. The healthcare allowance remains at zero because CZ does not register the treaty policy as Dutch insurance. The rent allowance is partially effective, but the assessed income differs due to the foreign income. The child-related budget only came into effect after the SVB link was established — and was paid out retrospectively.
What the results of the benefit assessments do not show is what the citizen has to do to achieve those results. And what goes wrong when they do not do so, or are unable to do so.
When filing an income tax return, the Dutch Tax and Customs Administration asks for your worldwide income: all income that you have earned anywhere in the world. For a cross-border worker, this means: your Belgian salary, your Belgian benefits, your Belgian allowances – everything.
However, citizens do not receive this information automatically. The Belgian tax authorities do not send it to the Dutch tax authorities. The DAC (Directive on Administrative Cooperation) infrastructure is designed for the automatic exchange of income data — but the exchange is not in real time and is not always complete. Citizens must enter the amounts themselves, based on their own payslips, benefit statements and their estimate of their Belgian income.
The Tax and Customs Administration explicitly places the responsibility on the taxpayer. Whatever the taxpayer enters is assumed to be true. If it is incorrect, a correction will follow later. And in the event of a correction, any overpayment may be reclaimed.
Figure XIII.2 — World income versus joint income. Belgian income is taxed in Belgium under the tax treaty and is exempt in the Netherlands. However, for the purposes of the allowances, it counts in full as global income. A euro in Belgium is not the same as a euro in the Netherlands — different purchasing power, different social security provisions, different cost structures. The allowance model does not take this difference into account.
The patient then tries to claim the medical expenses as a personal allowance. In the Netherlands, medical expenses above a certain threshold are deductible from income. This reduces the assessed income — and that, in turn, reduces the basis for calculating the benefits.
The strategy makes sense: deducting medical expenses reduces the assessed income, and increases the benefits. It is an indirect way of compensating the family for the costs that the system does not cover.
Figure XIII.3 — Medical expenses. The patient enters the medical expenses. The calculation model applies the deduction threshold. The result is zero. The assessment income does not fall, the benefits remain the same, and the actual costs continue to weigh on the family.
The Tax and Customs Administration consistently states that the responsibility for the accuracy of the tax return lies with the taxpayer. The taxpayer must:
And if the automatic exchange of information has not worked — if Belgium has not sent the data, or if the Netherlands has not processed it — that is no reason to delay filing the tax return. The taxpayer must sort it out themselves.
The Tax and Customs Administration places the responsibility for accuracy on the taxpayer. However, taxpayers cannot always obtain the correct information, as the automatic exchange of data between Belgium and the Netherlands is not comprehensive. The system shifts the risk of an administrative error onto those least able to cope with it.
Three allowances, three outcomes. One completely blocked, one partially blocked, one working. And the tax return itself asks the taxpayer to enter an income that they cannot calculate, placing the responsibility for its accuracy on them, with the consequences of a recalculation that automatically follows as soon as the Tax and Customs Administration has the actual figures.
What becomes apparent here is a pattern that recurs throughout the case file. The system has the tools, the laws and the intentions to support citizens. But the system does not interpret reality in the same way as citizens experience it. The benefits are calculated on the basis of an assessed income that treats a Belgian euro as equivalent to a Dutch euro. Medical expenses are entered but do not result in any deduction. And the citizen is personally responsible for the accuracy of an income figure of which they are unaware.
All organisations in this chapter are legal entities. Natural persons are identified by their role or their symbol (♂, ♀). The facts are based on Document B. This chapter does not constitute tax or legal advice. For the exact amounts and conditions, please consult the tax authorities or a tax adviser.
[[PAGEBREAK]]
The PD U1 is the European certificate of a person’s employment history. A cross-border worker needs it in both directions: to have their Belgian years recognised in the Netherlands, and to have their Dutch years recognised in Belgium. In both directions, the transfer is carried out manually. And the ESSPASS will not resolve this by 2030, as the individual remains the link in the chain.
Figure VIII.1 — Two directions. Direction 1 works: the NEO supplies the PD U1 for the Belgian employment history to the UWV. Direction 2 is blocked: the UWV supplies the PD U1 for the Dutch employment history, but the NEO refuses to integrate foreign employment history data into its system. In both directions, the citizen must personally take the document from one authority to the other.
The individual has worked in Belgium for 21 years. In order to claim Dutch benefits, that period of employment in Belgium must be verifiable within the Dutch system. To this end, NEO provides the PD U1 form to the UWV.
NEO provides three PD U1 forms. Upon comparison, it appears that five periods are missing — student work from 1999, gaps in 2001–2002, and transition periods. The citizen applies for the additional certificates themselves. The file is only complete once the information has been added manually. The system works, but only because the citizen detects and fills in the gaps themselves.
The individual also worked in the Netherlands for eight months. In order to claim Belgian entitlements, this period of employment in the Netherlands must be verifiable within the Belgian system. The UWV issues the PD U1 form for this purpose. However, NEO refuses to integrate foreign employment records into its system.
The UWV provides a PD U1 for the eight Dutch months. NEO refuses to integrate this data into its system: the agency regards itself as a sender, not a recipient. The only solution is for the citizen to manually enter the Dutch career data into the portal of another Belgian federal agency (FPS Social Security), which then transfers the data to MyCareer.
| Aspect | Direction 1 — BE → NL | Direction 2 — NL → BE |
|---|---|---|
| Publishing body | NEO (Antwerp) | UWV |
| Receiving body | UWV | NEO |
| Length of career | 21 years and 2 months | 8 months |
| Application | Manually, by email or via a form | Manually, using a form |
| Delivery | Works, with gaps | It works, but the recipient refuses |
| Transfer | Citizens must submit their applications to the UWV themselves | Citizen administers injection to FOD SZ |
| Blockade | 5 periods are missing; the citizen must fill them in themselves | NEO refuses to accept data from abroad |
| Bypass | Requesting additional certificates from employers | Manual entry via the FOD SZ portal |
In both directions, the citizen is the only party who submits the document, either physically or digitally. NEO and the UWV do send documents to each other, but do not receive them from one another. There is no automatic system-to-system link. In direction 1, the citizen fills in the gaps themselves. In direction 2, they bypass the uncooperative NEO via another federal agency.
Figure VIII.2 — ESSPASS in 2030. The wallet digitises the document. Verification becomes faster. But the citizen’s process remains the same: they apply for the document and share it with the verifying authority. Member States do not automatically link their systems together. In 2030, NEO will still refuse to integrate foreign employment data, and citizens will still have to use the same workaround.
| Aspect | Now (2026) | ESSPASS (2030) |
|---|---|---|
| Carrier | Paper | EUDI Wallet |
| Application | Manual, 75% no online application | Manually via the wallet |
| Issue | Manually, taking days to weeks | Automatically from the source, seconds |
| Transfer | In person, by post or by email | Manually via the wallet |
| Verification | Manually by a civil servant | Real-time QR scan |
| Fraud detection | Limited | Cryptographically strong |
| Coordination between Member States | None | None |
ESSPASS eliminates paper-based delays. It does not eliminate the need for manual synchronisation. The citizen remains the party who requests and shares the document. What was a physical action in 2026 will become a digital action in 2030. The citizen’s role does not change.
In Belgium, the main administrative obstacles to this case have now been identified and, in part, overcome via the tax bypass. What remains in Belgium are implementation issues — the reopening of the case, the disciplinary complaint and the labour audit. No new structural obstacles are emerging.
The situation is different on the Dutch side. There, the WIA, the SVB, the Tax and Customs Administration, the municipality of Hulst, the CAK, the RIV registers and the entire Dutch administrative apparatus are still awaiting their own analysis. And that is precisely where a structural difference with Belgium lies, one that has not yet been highlighted.
The manual U1 issue is the final Belgian obstacle in this matter. At the same time, it is the gateway to the Dutch part — because direction 2 (the Netherlands to Belgium) is the first step that also affects the Dutch authorities. What is still touched upon in passing in this chapter will take centre stage in the following chapters: the WIA, the SVB, benefits, healthcare allowance and local authority implementation.
The information on the PD U1 is based on Document B and on the official procedures of NEO and the UWV. The ESSPASS information is based on public sources from the European Commission and the European Parliament concerning the ESSPASS Regulation, the EUDI Wallet and the phased implementation of the Labour Mobility Package. This chapter does not constitute legal advice.
[[PAGEBREAK]]
Not everything that happens between Belgium and the Netherlands amounts to a deadlock. This chapter highlights one case where European coordination actually works: child benefit for a child who lives in the Netherlands and attends school in Flanders. Two authorities, two countries, one child. They manage to find common ground. And then a recalculation takes place.
For a child whose parents work in two different countries, the European Co-ordination scheme allocates child benefit. The rule is that one country pays the main contribution, and the other country tops up the difference to the level applicable in the country of residence. For this family, this means:
Figure XII.1 — The split. Parentia pays the Belgian contribution in the mother’s name (country-of-employment principle). The SVB tops this up to the Dutch level in the father’s name (country-of-residence principle). The child receives the full amount to which a Dutch child is entitled, divided between the two countries.
What works here is precisely what goes wrong elsewhere in this case. Two national authorities, two different sets of laws, two languages, and one child. They manage to find common ground.
The European Coordination Regulation provides for a mechanism known as the ‘overlap of family benefits’. This mechanism determines which country pays the main contribution and which country pays the supplementary amount. Parentia and the SVB apply this mechanism. They exchange information on each other’s cases, recognise each other’s decisions, and settle the amounts.
The key: the supplementary benefit in the father’s name. The mother worked in Belgium, so the SVB could not award a supplement in her name — Belgium was already responsible for that. However, the father did not work outside the Netherlands and lived in the Netherlands. This made him the appropriate person to receive the difference. A straightforward application of the European rules.
The European coordination of family benefits works here because both countries recognise each other’s systems and offset the amounts. Elsewhere in this dossier, things go wrong because no offsetting takes place — in the case of medical expenses, transport costs and allowances.
Now comes the second part of the story. As soon as the father’s retroactive reclassification to ‘disabled dependant’ status takes effect, the family’s tax situation will change.
What will change: the mother will have more money left over after tax. Not because she earns more gross income, but because the family will be entitled to tax relief following the recognition of the father’s disability.
And that’s when the cascade effect begins.
Figure XII.2 — The recalculation. The retroactive change in status to ‘dependent disabled person’ results in tax benefits for the family. On paper, the family’s income increases. Parentia reduces the contribution. The SVB recalculates the supplement. This may involve a recovery of benefits. What works today may be reversed tomorrow by an automatic recalculation.
This is the only chapter in Document C in which a mechanism works in its entirety. And that is precisely what makes it interesting, for three reasons:
Firstly, European coordination does indeed work for family benefits, because the regulation prescribes an explicit mechanism (overlap) and both countries adhere to it. Where that mechanism is lacking — such as in the case of medical conditions, healthcare allowances and patient transport — it does not work.
Secondly, the fact that the supplement is issued in the father’s name shows that the same administrative reality yields different results depending on who submits the application. If submitted in the mother’s name: no supplement. If submitted in the father’s name: a supplement is issued. One family, two outcomes, depending on the formal role.
Thirdly, the retroactive reclassification poses a new risk. As soon as the father’s status changes, the amounts are recalculated. What is currently a functioning system could result in a debt next year – not because of an error, but because of a recalculation.
These three points make this chapter the perfect counterpart to the Parentia case (Block II) and the Hulst Local Authority case (Block III). The same authorities, a different outcome. What this case study shows is that the system sometimes works — but it is never guaranteed. Success is the exception, not the rule.
The facts are based on Document B. The retrospective recalculation is a forecast based on the pattern of recalculations in the file; the final amounts and the outcome have not yet been determined. All organisations in this chapter are legal entities. Natural persons are identified by their position or their symbol (♂, ♀).
[[PAGEBREAK]]
A cross-border worker lives in the Netherlands and receives treatment in Belgium. His entire medical record is stored in Belgian data silos. What happens if something goes wrong in his country of residence? An emergency, a fall, an acute complication? Who, then, is the doctor who knows him? The answer is: no one. The law prohibits it.
In the Netherlands, a patient may only be registered with one GP practice. Anyone who registers with a new practice is automatically deregistered from their old one. This is a nationwide system designed to prevent duplicate registrations. For the average Dutch person, this makes sense. For a cross-border worker who has spent their entire medical history abroad, however, it is a problem.
The reason why dual registration is prohibited: the Dutch Health Insurance Act and the administrative regulations governing general practice prohibit dual registration across national borders. A GP may only register a patient if they actually bear primary and active medical responsibility for that patient. Registration without active treatment is not legally possible.
Figure XIX.1 — The legal impossibility. The patient’s complete medical records are held in Belgium. He cannot register with a GP in the Netherlands, as the law prohibits dual registration. He lives in a country where no doctor is familiar with his medical records.
What happens if a medical problem arises in the country of residence?
The file explicitly sought to resolve this issue by registering the patient with a Dutch GP practice for purely informational purposes. No active treatment was provided; the aim was simply to ensure that the patient’s medical records were accessible on Dutch territory, so that an on-call doctor or an emergency doctor could see what the situation was.
That application was rejected. The practice acknowledged over the phone that, in an ideal world, this would be the right solution, but that the law does not permit it. Registration without an active treatment relationship is not legally possible. The letter confirmed the system block, leaving no room for reconsideration.
Figure XIX.2 — The practice’s response. The GP practice acknowledged that, in theory, an ‘informative’ registration would be the ideal solution, but that the law prohibits dual registration across borders. The practice cannot register a patient without an active treatment relationship.
The rule that a patient may only be registered with one GP practice is intended to safeguard the quality and continuity of care. For ordinary citizens, this rule works. For cross-border workers who have spent their entire medical history in another country, however, this same rule creates a gap in their healthcare security.
The European Union recognises the right to free movement. It also recognises the right to medical care abroad. However, it does not provide a solution for patients whose medical records are held in one Member State whilst they are physically in another. The law safeguards continuity within a single country. It does not provide for continuity across borders.
What the family is left with is a pragmatic solution: the patient carries his own medical history with him, either in physical or digital form. If he has to see a Dutch doctor in an emergency, he must be able to explain the situation himself. If they are admitted to hospital in the Netherlands, they must provide the Belgian medical records themselves. They are not only the data hub for the administration, but also for their own immediate medical safety.
Figure XIX.3 — The paradox. The law safeguards continuity of care within a single country. It does not provide for continuity across borders. For the average citizen, the rule works. For a cross-border worker, that same rule creates a gap in protection.
The patient lives in the Netherlands, but his entire medical history is in Belgium. Dutch law prohibits patients from being registered with two GPs. The patient is therefore unable to establish a medical base in the country where he lives. And in an emergency, there is no doctor who is familiar with his medical records.
The GP practice approached about this acknowledged that, in an ideal world, this would be the right solution. The law prohibits it. The European Union recognises freedom of movement and cross-border healthcare, but does not provide for continuity of care across borders. The family gets round this by having the patient carry their own medical records. Once again: not because the system requires it, but because there is no alternative.
All organisations mentioned in this chapter are legal entities. Natural persons are referred to by their title or their symbol (♂). The facts are based on Document B. This chapter does not constitute legal or medical advice.
[[PAGEBREAK]]
All the previous chapters have described authorities that refuse to act, procedures that grind to a halt and systems that fail to recognise the reality of the situation. This chapter shows the result. When benefits fail to materialise, allowances remain at zero, medical costs continue to mount up and no one steps in to bridge the gap, there is only one player left: the private sector. The bank. The energy supplier. The network operator. And they do not operate according to protocols, but according to payment deadlines.
Figure XX.1 — The bank account as a reserve. Any shortfall in benefits and allowances, combined with outgoings on medical costs, creates a deficit in the bank account. That deficit is covered by commercial parties: the bank, the energy supplier, the debt collection agency. In return for interest, subject to penalties, and under strict conditions.
When government payments stop and medical bills keep mounting up, the joint bank account goes into the red. The credit card reaches its limit. ING offers a payment plan: the entire debt is transferred to a separate account and repaid over twenty-four months.
The bank’s conditions: no overdrafts for three months. No credit card for twelve months. Monthly repayment of €286.55. Anyone who fails to comply with the conditions will lose their place on the scheme.
The bank is not acting in bad faith here. It is simply applying commercial logic. But the logic of the market is not the same as that of social justice. The public authorities that caused the shortfall bear no responsibility for the interest and penalties the family has to pay.
The energy supplier is seeing payment arrears mount up. Payment arrangements are agreed, broken and reinstated. The monthly instalments are increased to prevent a back payment. And when the patient invokes the ‘medically vulnerable consumer’ protection status, Vattenfall requires a statement from an independent Dutch doctor.
That requirement is impossible for a cross-border worker to meet. GPs refuse, the GGD refuses, and the private sector charges exorbitant fees for a remote medical examination. The family submits a Belgian medical certificate. Vattenfall refuses to recognise it. Only after months does the supplier admit that the requirement was unjustified and that the authority lies with the network operator.
Ultimately, the outstanding debt is referred to the debt collection agency Flanderijn. The total debt amounts to over €2,300.
Following the late referral from Vattenfall, the file is passed on to the network operator. Stedin is legally responsible for the physical supply. The patient must resubmit all medical evidence to a new private provider. There is no data transfer whatsoever, nor is there any recognition of documents previously submitted.
Stedin is the organisation that can guarantee the electricity supply, which is crucial for the refrigerated storage of insulin. However, the same applies here: the patient must rebuild the entire case from scratch, without the previous application to Vattenfall being recognised.
Figure XX.2 — Who bears the blame? The government causes the shortfall but accepts no responsibility. The private market fills the gap and collects the interest. The family bears the consequences: loss of income, loss of benefits, mounting interest, collection costs, and further impoverishment.
The chain of failing public authorities ultimately leads to one central point: the bank account. What does not come in, and what does leak out, determines how far the family’s circumstances deteriorate. And what the family cannot cope with is absorbed by the commercial market — at interest, subject to penalties, and with collection charges.
Private sector organisations are not to blame here. They are simply applying commercial logic. What this chapter shows is that the government has shirked part of its responsibility. The gap it has left is being filled by the market. But the market does not fill gaps out of a sense of solidarity. The market fills gaps at a price.
And that price is paid by the family that already has no income. By the parents who already receive no benefits. By the patient whose amputations are not compensated by anyone. By the child who plays no part in this game but who bears the consequences.
All organisations mentioned in this chapter are legal entities. Natural persons are referred to by their position or their symbol (♂, ♀). The facts and figures are based on Document B. This chapter does not constitute legal or financial advice.
[[PAGEBREAK]]
When national avenues have been exhausted and national politicians fail to intervene, there is still one level left: Europe. In theory, Europe should provide the solution to a cross-border problem. In practice, however, it appears that most European bodies lack the mandate to break the deadlock in the national chain. But two exceptions have emerged in this case. Two opportunities that nobody had anticipated.
Figure XXII.1 — The European and international levels. Each with its own mandate, each with its own limitations. Two exceptions stand out: the European Parliament, where petition 1034/2026 has been made public and is open to co-signatories, and the Commission, which has been investigating several infringement complaints against Belgium and Flanders since 2022.
Figure XXII.2 — The circular reference. The European institutions refer to one another. The Commission is waiting for the Member States. The Member States are waiting for the Commission. SOLVIT is waiting for a clear infringement. The ELA is waiting for a request from the Member States. Parliament cannot review national decisions. The ECHR is waiting for national legal proceedings.
In September 2026, two openings appeared in this seemingly closed circle that nobody had anticipated.
The European Parliament has taken up petition 1034/2026. On 25 September 2026, the status was changed to ‘open to co-signatories’. This means that the petition has been given an official parliamentary summary and has been published on the European Parliament’s petitions portal.
The summary has been published under the policy areas of the Internal Market, Employment, Health, Social Affairs and Fundamental Rights, with Belgium and the Netherlands listed as the Member States concerned. The petition highlights the structural failure of both Member States to comply with the Vester judgment, their refusal to aggregate insurance periods, and the residence requirements. It calls on the European Commission to initiate infringement proceedings.
What this means is that the petition is no longer a private complaint sent to a PO box. It is a formal parliamentary dossier that any EU citizen can sign. This changes the political status of the dossier.
On 25 September 2026, an official letter from the European Commission (DG EMPL) was added to the file. It reveals that the Commission has been investigating several infringement complaints against Belgium and Flanders since 2022 and 2023. These complaints relate to the Flemish job bonus, student finance and the Growth Package — all three of which fall under Regulation 492/2011 on the free movement of workers.
The Commission has not reached a final decision in all these years. That does not mean that the complaints have been rejected. It means that they are still under consideration. And that this family’s complaint forms part of a series of complaints that the Commission is already investigating.
Figure XXII.3 — The two avenues. On the one hand, the petition in the European Parliament is public and open to co-signatories. On the other hand, it appears that the European Commission has been investigating several complaints against Belgium and Flanders since 2022 without reaching a decision. This family’s complaint is not an isolated case.
The European Commission has one binding instrument: the infringement procedure. Where a Member State systematically breaches EU law, the Commission may initiate proceedings which ultimately lead to a ruling by the Court of Justice and, in theory, to financial penalties for the Member State.
In practice, this is a political decision, not a legal formality. The Commission assesses whether it is politically expedient to initiate infringement proceedings. In this case, the complaint has been registered under CPLT(2026)01544 and merged with other complaints under Ares(2026)5891709. The Commission has not yet initiated infringement proceedings.
The reasons why this does not happen are structural:
What this means for the public: even if the Commission were to intervene, the proceedings would take years. And even if the Court were to rule that a violation had taken place, that would not immediately resolve the individual case.
Figure XXII.4 — The European paradox. The EU coordinates social security, but does not harmonise it. As long as this remains the case, any solution will depend on national governments.
The European level has created the framework within which cross-border social security should be organised. It has laid down the rules, established the liaison networks and set up the coordinating bodies. And it has entrusted itself with the task of monitoring compliance with the rules.
However, the European level has failed to ensure compliance with the architecture. Most bodies can only monitor, mediate and report — without any enforcement powers. And the path to the courts is a long one.
Two things stand out. The petition in the European Parliament has been made public and is open to co-signatories. This changes the political status of the case: it is no longer a private complaint. Furthermore, the European Commission has been investigating several infringement complaints against Belgium and Flanders since 2022. This family’s complaint is one of a series that is already underway.
What remains is a citizen who has knocked on every door in Europe and received the same message at every one: we have no jurisdiction over your individual case. But whilst some doors remained shut, two have been left ajar. For a case that went unheard for ten years, that is no small thing.
All organisations in this chapter are legal entities. The legal basis for the various procedures is set out in the Treaties of the European Union (Articles 258 and 259 TFEU for the infringement procedure), the Rules of Procedure of the European Parliament (PETI procedure) and the Statute of the Council of Europe (ECHR). The information on the complaints procedures is based on Document B. This chapter does not constitute legal advice.
[[PAGEBREAK]]
When an implementing body refuses, there is a level of authority that is expected to intervene: the ombudsmen, the mediation services, the European networks. In this case, that level was called upon. It listened, investigated, and concluded at every level that it could not intervene. In the end, it turns out that the most robust mediation route — the European Labour Authority — is not even accessible.
Figure XV.1 — Four mediators, four restrictions. The Federal Ombudsman may intervene in ongoing cases, but cannot compel action. The National Ombudsman cannot intervene whilst an administrative appeal is pending. SOLVIT cannot compel action and withdraws from ongoing proceedings. The GIP can only provide information and refuses to report a system error.
The Belgian and Dutch ombudsmen appear similar, but their legal frameworks are fundamentally different.
Figure XV.2 — The two ombudsmen. The Federal Ombudsman may intervene in ongoing administrative cases. The National Ombudsman is legally precluded from intervening whilst an administrative appeal is pending.
Within the European framework, SOLVIT is the body responsible for mediating when national public authorities breach EU law. Two SOLVIT centres are involved in this case: SOLVIT Netherlands and SOLVIT Belgium. What follows is a pattern of mutual referral — until both centres withdraw.
Figure XV.3 — The circular referral between SOLVIT NL and SOLVIT BE. Both centres refer cases to one another. And when the patient asks to be referred to the European Labour Authority, both refuse: SOLVIT NL because no infringement has been established, and SOLVIT BE because referral does not fall within its remit.
If SOLVIT is unable to provide a solution, there remains one European body that was set up specifically to deal with cross-border disputes: the European Labour Authority (ELA). The ELA was established to promote fair labour mobility and to facilitate mediation between Member States in cross-border social security disputes.
In theory, the ELA should be the body that steps in when national governments cannot reach an agreement. In practice, however, the ELA appears to be inaccessible to individual citizens.
Figure XV.4 — The ELA as a closed door. The ELA only acts as a mediator at the request of Member States. Direct requests from citizens are declared inadmissible. And the referral points that are supposed to pass on these requests (SOLVIT NL and BE) refuse to do so. The body set up specifically for this sort of situation remains out of reach for those who need it most.
What all mediators have in common is one fundamental characteristic: they can investigate, make recommendations and report, but they cannot compel. No mediator can oblige an implementing body to review a decision, award a benefit or process a registration.
There is a second characteristic to consider: each mediator has their own threshold for when they may intervene. The Federal Ombudsman may do so in ongoing cases. The National Ombudsman may not. SOLVIT may not do so in ongoing legal proceedings. The GIP may never do so. The ELA may only do so at the request of Member States. The ECHR may only do so after all domestic remedies have been exhausted.
And this dossier shows that the most robust mediation route — the European Labour Authority — is not only limited, but completely closed off. The ELA was designed for situations such as this, but is not accessible to individual citizens. And the only intermediaries that could grant access (SOLVIT NL and BE) refuse to do so.
The bottom line is that the mediation process acts as a series of filters. Every complaint is recorded, investigated and referred to another body which is also unable to take action. And when the final body is reached, it too turns out to have closed its doors.
All organisations mentioned in this chapter are legal entities. Natural persons are referred to by their job title. The facts are based on Document B. The legal frameworks are based on official communications from the relevant authorities. This chapter does not constitute legal advice.
[[PAGEBREAK]]
When no other body intervenes, the courts are the last resort. In theory, this is where a decision can be enforced. In practice, however, access to the courts themselves proves to be a barrier — and that barrier is precisely the problem the courts were meant to resolve.
Figure XVI.1 — The two legal routes. In Belgium, the labour court can be accessed by means of a simple application, without any court registration fee or litigation costs for the citizen. The only costs are the lawyer’s fees. In the Netherlands, the process begins with a notice of objection lodged with the authority itself, followed by an appeal to the administrative court, which involves court fees and lawyers’ fees.
Both countries have a system of subsidised legal aid: ‘pro deo’ in Belgium, and ‘toevoeging’ in the Netherlands. Anyone whose income falls below a certain threshold is assigned a solicitor, the costs of whom are (largely) borne by the state. However, the eligibility criteria pose a problem for a cross-border worker facing a medical emergency.
Figure XVI.2 — The pro bono thresholds. In Belgium, the assessment is based on the net monthly income from two years earlier (the reference year). In the Netherlands, the assessment is based on the gross annual income from two years earlier. In both countries, actual income has fallen to zero during the medical crisis, but the reference-year income is still based on the period before the crisis. On paper, the individual earns too much, whilst in reality they have nothing.
The crux of the problem is the reference year. Both countries assess entitlement to subsidised legal aid on the basis of income from two years earlier. This is intended to prevent fraud: anyone who rapidly reduces their income in order to qualify is barred from the scheme. However, the system does not take into account a sudden, involuntary drop in income due to illness.
In the user’s file:
The result is that, on paper, the individual earns too much to qualify for legal aid, whilst in reality they have no money to pay for a solicitor. The threshold, which is intended to protect the most vulnerable, actually excludes those who have fallen on hard times due to a sudden crisis.
Figure XVI.3 — The irony of the reference year. The system assesses income from two years ago. In the context of a medical crisis, that income is no longer representative. On paper, the citizen earns too much to qualify for legal aid, whilst their actual income is zero. The threshold, which is intended to protect the most vulnerable, actually excludes precisely those who have fallen on hard times due to a sudden crisis.
In theory, legal action is the ultimate solution. A court can compel an authority to review a decision. In practice, however, access to the courts themselves proves to be a barrier. And that barrier is precisely the problem the court was supposed to resolve.
A citizen who cannot afford a solicitor cannot take legal action. A citizen who is not eligible for legal aid because their reference-year income is too high cannot take legal action. Citizens who have no income because their benefits have been suspended cannot take legal action. The system refers people to the courts, but the courts are out of reach for those who need access to them most.
In Belgium, the threshold is lower: the application form is straightforward and free of charge, and the only costs involved are the lawyer’s fees. But even those fees are unaffordable for many people. In the Netherlands, the barrier is higher: court fees, lawyers’ fees, and an even stricter assessment framework.
All organisations mentioned in this chapter are legal entities. The income thresholds for pro bono and legal aid are based on official communications from the relevant authorities (advocaat.be, the Legal Aid Board). This chapter does not constitute legal advice. If in doubt, consult a specialist service or a solicitor.
[[PAGEBREAK]]
Politics has the power to change laws, allocate budgets and steer implementing bodies. In theory, it is the ultimate corrective power. In practice, however, it appears that politicians only exercise that power when the problem is presented as a collective systemic failure — not when it is raised as an individual complaint.
Figure XVII.1 — The separation of powers. Belgium has a classical trias politica, enshrined in Articles 33, 36, 37 and 40 of the Constitution. The Netherlands does not have a strict separation of powers, but rather a system of checks and balances, centred on ministerial responsibility (Article 42) and the duty to provide information (Article 68).
Figure XVII.2 — What politics can and cannot do. Politics can make and amend laws, but cannot review individual decisions. It can issue new instructions to an implementing body, but cannot open or close an individual case. The separation of powers prohibits interference in specific cases.
In Belgium, the Chamber of Representatives has a wide range of oversight tools at its disposal. The standing committees — including the Committee on Social Affairs, Employment and Pensions — draft legislation and scrutinise the government. Within these committees, Members of Parliament may ask written and oral questions, and raise interpellations which may lead to motions of recommendation or no confidence.
In addition, there is the Petitions Committee. Any citizen may submit a petition to it. The committee assesses whether the petition is admissible and then forwards it to the relevant committee or the relevant minister. That minister is required to provide a written response within six weeks. The committee may then decide to close the case, refer it elsewhere, or organise a hearing.
Figure XVII.3 — The Belgian way. A member of the public submits a petition. The Petitions Committee declares it admissible and forwards it to the relevant committee and the minister. The minister provides an explanation within six weeks. That is where the process ends. In this case: declared admissible, forwarded, no substantive response.
The House of Representatives has a similar range of instruments at its disposal. Members of Parliament may table written questions, ask oral questions, raise interpellations and table motions. The Standing Committee on Social Affairs and Employment (SZW) is responsible for scrutinising the Ministry of Social Affairs and Employment and the implementing bodies under its remit.
What sets the Netherlands apart is the parliamentary inquiry. This is the most powerful investigative tool: a committee that can hear witnesses under oath, request documents and publish a public report. The benefits scandal is the most recent example of how this instrument works — and how long it takes before it is deployed.
Figure XVII.4 — The Dutch approach. The House of Representatives does not deal with individual complaints. The Social Affairs and Employment Committee may shelve a case. The most powerful tool — the parliamentary inquiry — requires a majority and is only used in the event of major systemic crises. The benefits scandal illustrates just how long that takes.
| Aspect | Belgium | The Netherlands |
|---|---|---|
| Individual complaint | Via the Petitions Committee · admissibility assessment · referral to the minister · written explanation within 6 weeks | Not under consideration · legal avenues must remain open or have been exhausted · a letter to an MP rarely leads to action |
| Parliamentary instrument | Written and oral questions · interpellations · motions for a recommendation · motion of no confidence · committee of inquiry | Written and oral questions · interpellations · motions · thirty-member debate · parliamentary inquiry |
| Strongest remedy | Parliamentary Committee of Inquiry · right of inquiry (Article 56 of the Constitution) | Parliamentary inquiry · witnesses under oath · legal basis (Parliamentary Inquiry Act 2008) |
| Case study in this dossier | The petition was declared admissible, forwarded, and then there was silence. Letters were sent to nine parties; five did not respond, and four referred the matter to their research department or stated they were not authorised to respond. | Urgent letter registered; set aside pending committee debate; no individual action taken. Parties were referred to the Ombudsman or the research department. |
| Lead time | Minister: 6 weeks for a written explanation · followed by a decision by the committee | No deadline for dealing with individual letters · survey: years |
This is the key insight. The same facts can be presented in two different ways, and those two presentations elicit two completely different political responses.
Figure XVII.5 — Two presentations, two responses. The same facts, presented differently. As a personal complaint: politicians declare themselves to have no jurisdiction. As an institutional audit: politicians have no choice but to address a systemic problem.
This report is not a complaint. It is an audit. That is not a rhetorical choice — it is a strategic choice.
A complaint is treated as an individual problem. An audit is treated as a systemic problem. And only a systemic problem triggers the political mechanisms that can lead to real change.
The documents in this file are therefore organised in the form of an audit:
The difference is not merely superficial. It determines whether politicians can brush the matter aside — “we do not deal with individual cases” — or whether they must address it.
All organisations in this chapter are legal entities. Natural persons are identified by their position or their symbol (♂, ♀). The facts are based on Document B. The constitutional basis is Articles 33, 36, 37 and 40 of the Belgian Constitution, and Articles 42 and 68 of the Dutch Constitution. The parliamentary procedures are based on the rules of procedure of the Chamber of Representatives and the House of Representatives. The information on the benefits scandal is based on public reports by the National Ombudsman and the parliamentary committee of inquiry. This chapter does not constitute legal advice.
[[PAGEBREAK]]
In Belgium and the Netherlands, the King is the symbolic head of state. Formally, he has no independent power: his actions are only valid if co-signed by a responsible minister. Nevertheless, he receives thousands of petitions from citizens every year. These petitions are not legal proceedings. They are an informal channel — and precisely because they are informal, they can influence the political agenda in a way that formal complaints cannot.
Figure XVIII.1 — The formal position of the King. In Belgium, the King is part of the legislature (Art. 36) and holds executive power (Art. 37), but no act is valid without the co-signature of a minister (Art. 106). In the Netherlands, the King is inviolable (Art. 42) and does not possess any independent constitutional powers. In both countries: the King reigns but does not govern.
In both countries, the right to submit a petition is enshrined in the constitution.
Figure XVIII.2 — The right to petition. In Belgium, the right to petition is enshrined in Article 28 of the Constitution, and Article 57 grants the Chamber of Representatives the power to refer petitions to ministers, who are obliged to provide an explanation. In the Netherlands, the right to petition is enshrined in Article 5 of the Constitution, without any specific parliamentary procedure. The House of Representatives has its own regulations governing the Committee on Petitions and Citizens’ Initiatives.
Figure XVIII.3 — The path of a petition. In Belgium, the Palace’s Petitions Service receives around ten thousand petitions each year. Each petition is examined and either given a response or referred to the most appropriate authority. In the Netherlands, the King’s Cabinet receives between three thousand and five thousand letters each year. The Cabinet analyses these and forwards them to the relevant minister. In both cases: there are no binding consequences.
Formally speaking, the King has no power. The minister is responsible. Politics calls the shots. And yet: a petition to the King reaches a level that an ordinary complaint does not.
Figure XVIII.4 — The paradox of the informal channel. An ordinary complaint remains at the executive level. A petition to the King reaches the political level, because it is analysed by the Cabinet and must be answered by a minister. The King has no power, but his postbox does.
In this case, a petition has been submitted to the head of state in both countries.
Figure XVIII.5 — The two petitions. In both countries, a petition was sent to the Head of State. In Belgium, the digital channel failed due to a DNS error on the Palace’s mail server; only the physical letter reached the Palace. In the Netherlands, the digital channel did work. In both cases, the petitions were forwarded to the relevant minister, without any substantive intervention by the King.
The nobility — the King and his cabinet — have no executive power in Belgium and the Netherlands. The King reigns, but does not govern. He is immune from prosecution, but his acts are only valid if countersigned by a responsible minister.
Nevertheless, a petition to the King is a channel that does reach the political level. An ordinary complaint remains within the executive branch. A petition to the King is analysed by the Cabinet, answered by a minister, and — more importantly — the minister bears political responsibility for it. In this way, the message reaches a level where the executive branch cannot ignore it.
The paradox is that the King has no power, but his postbox does. It is an informal channel. No legal procedure. No right to a decision. But it is a channel that takes the message one level higher. And in a case where those in charge of implementation refuse to listen, that is no small matter.
All organisations in this chapter are legal entities. Natural persons are referred to by their title or symbol. The constitutional basis is Articles 28, 33, 36, 37, 57 and 106 of the Belgian Constitution, and Articles 5, 42 and 68 of the Dutch Constitution. The procedure for petitions is based on the official communications from the Royal Palace and the King’s Cabinet. This chapter does not constitute legal advice.
[[PAGEBREAK]]
Alongside the government and the private sector, there are organisations that are supposed to support citizens: trade unions, equal opportunities centres, human rights organisations and academics. This report looks at four of them. One actually helps, with a tax service that makes a real difference. One claims it has no jurisdiction on the basis of a technical criterion. And the academic world only gets involved once the case has become irrefutable — and then things move quickly.
The General Christian Trade Union is Belgium’s largest trade union. As well as providing trade union representation, it offers legal assistance, tax advice and payment services for unemployment benefits. In this case, the trade union plays three roles.
The annual Belgian tax return is prepared via the ACV East Flanders tax department. The disability code is entered correctly. The childcare certificates are added manually. The result is a refund of €2,669.97, which was definitively submitted on 25 September 2026. This service is functioning correctly.
If the CM refuses to pay sickness benefit and does not apply the aggregation rules, the trade union is formally called upon to intervene. It is not the patient who is a member, but his partner — herself a trade union representative in the education sector. Through her role, she seeks assistance from the Christian Teachers’ Association, the education branch of the same trade union.
The legal department fully acknowledges the system error. The administrative distinction between incapacity and disability is confirmed. Nevertheless, the trade union refuses to take legal action against the CM, citing a formal argument: the patient herself does not have a membership number. The paying member — the partner — suffers the financial consequences, but receives no legal assistance for a case that is not in her name.
Figure XXI.1 — ACV: two entry points, two outcomes. The tax department is functioning. The legal department stops at the individual membership threshold: only members who have a case file of their own receive assistance.
UNIA is the Interfederal Centre for Equal Opportunities. It is Belgium’s independent watchdog against discrimination. When administrative bodies refuse to recognise a citizen on the grounds that they reside in a neighbouring country, the question arises as to whether this constitutes discrimination.
In April 2026, UNIA will be formally notified of the systematic exclusion of cross-border workers. The notification concerns the refusal to recognise medical qualifications solely on the grounds of residence in the Netherlands.
UNIA’s assessment highlights a shortcoming in Belgian anti-discrimination legislation. The watchdog formally declares that it lacks jurisdiction, for two critical reasons.
Figure XXI.2 — UNIA’s lack of jurisdiction. UNIA declares that it lacks jurisdiction for two reasons: place of residence is not a protected ground under anti-discrimination legislation, and the exclusion stems from the Act itself.
UNIA is closing the case. In doing so, it unwittingly confirms that the exclusion of cross-border workers under the current system is entirely legal.
The academic world is the final line of defence that could come to the public’s aid. Universities and centres of expertise study systemic failure, publish research on institutional exclusion, and possess the analytical tools to evaluate a case scientifically. What happens when they are presented with a real-life case?
Between May and September 2026, various academics and institutions will be contacted. Maastricht University’s ITEM Centre of Expertise, the Institute for Social Law at KU Leuven, the Montaigne Centre at Utrecht University, the DRIFT Transition Institute at Erasmus University, and Amsterdam UMC. For months, there is no response. Not a single substantive response is received to a dossier that embodies precisely the themes on which these institutions publish.
When, in the autumn of 2026, the complete dossier exceeds the threshold of 250 pages of irrefutable empirical evidence, the ITEM centre of expertise will begin a substantive review. On 17 September 2026, a professor confirms receipt of the dossier and states that he will read the documents first before commenting on their content.
In the following week, the case is actively reviewed. An independent researcher specialising in cross-border employment — based at an academic institution and not involved in the case — examines the material and, within a few days, provides a series of legal analyses and alternative courses of action:
In addition, the researcher receives formal written confirmation from the FPS Finance (Non-Residents Tax PG15 Ledeberg) that voluntarily attaching medical certificates to the tax return is a valid way of proving tax disability status. A few days later, he submits an official letter from the European Commission (DG EMPL, 25 September 2026), which shows that the Commission has been investigating three infringement complaints against Belgium and Flanders since 2022 and 2023 without reaching a decision.
What seemed impossible for months has happened in four days. The academic community, which had previously remained silent, has delivered a series of legal breakthroughs in the space of a week. The academic world operates retrospectively, not curatively — and once it gets going, it moves quickly.
Figure XXI.3 — From silence to disclosure. For months, the academic world failed to respond to the dossier. It was only when it contained 250 pages of irrefutable evidence that the ITEM centre of expertise agreed to conduct a substantive review. And then things moved quickly: within four days, new legal avenues, court rulings and system links were provided.
The four allies who are supposed to assist the citizen all operate within their own borders.
The trade union works on behalf of its members, but only in relation to the case filed in their name.
UNIA campaigns against discrimination, but only against discrimination that does not stem from the law itself.
The academic world operates on the basis of theory, but only comes into its own once the case has become complete and irrefutable. What had remained dormant for months springs into action within a matter of days. The focus is then no longer on the case itself, but on its scientific validation.
And what remains for the public is that the structural exclusion of cross-border workers has not only been legalised, but has also become socially normalised. Those who build up a case and persevere will eventually be heard. But only those who persevere.
All organisations mentioned in this chapter are legal entities. Natural persons are referred to by their position. The facts are based on Document B. UNIA’s role and mandate are based on Belgian anti-discrimination legislation (Act of 10 May 2007). The judgments cited are C-279/93 (Schumacker), C-134/18 (Vester) and C-385/17 (Hendrix). This chapter does not constitute legal advice.
[[PAGEBREAK]]
Five layers. Twenty-three chapters. One body at the centre. This is the complete System Audit board, at a glance.
Figure XXIII.1 — The complete pyramid. Five layers. At the top: Europe and the international sphere; then symbolism; then politics; then oversight and the judiciary; and at the base: implementation and the private market. At the very bottom: the individual and the family, for whom everything is ultimately done.
Five layers, each with its own logic, each with its own mandate, each with its own limitations. What the case has shown is not that one layer is failing, but that the layers are failing together in a way that nobody can fix.
What remains is the body. A body that has been viewed, described and assessed through every layer. And which has never been fully recognised as a single whole. The individual themselves acts as a data hub, a repository of records, a connecting cable between two systems that are incompatible with one another. And when they can no longer fulfil that role, the system grinds to a halt.
Document C has outlined the framework and the mechanisms. Document D draws the conclusions. Three building blocks for reform, based on the realisation that the fault does not lie with the individual civil servant, but in the very architecture of the coordination process itself.
What has become apparent in Document C is the structure of the vacuum. What is set out in Document D is the structure of the solution. Not because a single measure solves everything, but because the architecture can be adapted at three points where it is currently stalling.
This is the conclusion of Document C. All organisations mentioned in this document are legal entities. Natural persons are identified by their role or their symbol (♂, ♀). Document C is the visual counterpart to Document B (data matrix) and Document A (narrative audit). The conclusions and building blocks are set out in Document D (synthesis).
[[PAGEBREAK]]
Each of the previous chapters illustrated a single mechanism, a single instance or a single pattern. This chapter shows everything at once. The same nodes, the same channels, the same colours — but now in a single image, in no particular order, just as the family experiences it.
Readers who have got this far will have become familiar with each component individually. Each organisation had its own chapter. Each mechanism had its own diagram. Each colour had its own meaning. But the family facing this situation does not have that sense of order. The family sees only the whole picture. And that whole picture is a jumble.
This is the whole thing.
Figure XXIV.1 — The complete network. What the family sees when it tries to navigate. Two worlds, each with its own set of organisations. One family in the middle. And 127 connections, most of which are blocked. In the previous chapters, each component could be followed separately. Here is the whole picture — without order, without hierarchy, without tranquillity.
All the previous chapters were well-organised. Each had a title, a framework and a clearly defined subject. Anyone reading those chapters gets the impression that the system is comprehensible, that each part has its own function, and that the chaos is merely an illusion.
This diagram dispels that impression. There is no structure here. There is no title for each node. There is only the whole: 54 organisations, one family, 127 connections, 96 of which are blocked. And in the centre, a body that must hold its own.
Readers who have worked their way through the previous 23 chapters knew where to look. They were able to follow the connections, recognise the channels and interpret the colours. But the family facing this situation does not have that structure. They do not realise that Vitaz operates in the healthcare sector and CM in the financial sector. They do not realise that CZ administers a treaty policy and CAK manages the register. All they see is this: there are people who are supposed to process my file, and they are not doing so.
That is what this diagram illustrates. Not the failure of a single body. Not the failure of a single mechanism. But the failure of the whole system — a network so complex that it is impossible for an ordinary person to find their way through it, and so rigid that it is impossible for an ordinary person to change anything within it.
The 127 connections shown in this diagram are the most important ones. In reality, there are more. Every letter sent to an organisation, every time a civil servant looks at a system, every telephone call made — these are all connections that do not appear in this diagram.
And the 96 blocked connections shown here are only the blocked connections between the family and the organisations. In reality, there are also blocked connections between the organisations themselves. Between CM and UWV. Between CZ and CAK. Between FOD SZ and FPS Finance. Every blocked connection between organisations is an extra layer that the family has to navigate.
What this diagram shows is not the entire network. It is a representation of the entire network. If all the connections were drawn in, the diagram would become illegible. And that is precisely the point.
All organisations in this chapter are legal entities. Natural persons are referred to by their title or symbol. The organisations are based on Document B. The connections are based on the actual correspondence in Document B, summarised where relevant. The figures (54 organisations, 127 links) are indicative; the actual number is higher. This chapter does not constitute legal advice.
↑