Study no. 4 · Sovereignty and asset data

Delegated operation: the works come back, the capability does not

Nizar Younes Mqam ORCID 0009-0008-5549-7620

Study · 19 September 2026 · Cahiers de technopolitique, no. 4

French law is generous to the public owner: the works needed for the service belong to it from the moment they are built, even if the contract says nothing. A good concession contract goes further and guarantees their condition at handback. But neither the law nor the contract returns what makes it possible to use them. The capability to operate is neither inventoried, nor valued, nor required. It leaves with the outgoing operator, and it has to be bought back.

What the law returns, and it returns a great deal

Let us start with what works, because it is often forgotten when delegation is discussed. French law protects the public owner far better than is generally believed, and it does so without asking anything of it.

On 21 December 2012, the full judicial assembly of the Conseil d'État, the French supreme administrative court, sitting in its most solemn formation, handed down in the Commune de Douai case a decision of principle on the fate of the assets of a concession. The ground that matters fits in one sentence (my translation):

In the context of a public service delegation or a works concession placing on the contractor the investments corresponding to the creation or acquisition of the assets necessary for the operation of the public service, all of these assets, movable or immovable, belong, where the agreement is silent, to the public person from the moment they are built or acquired.

The force of this rule must be measured. It does not say that the assets will come back at the end. It says that they belong to the public person from the moment they are built, and that the silence of the contract changes nothing. A delegated operator that builds a treatment plant, lays a network or installs equipment builds, at the very second it does so, an estate that is not its own. The court adds that these works fall under the regime of the public domain as soon as they stand on the property of a public person, and that any derogation, as regards real rights granted to the contractor, remains bounded by the limits set by statute and by one condition: that the nature and use of the rights granted must not be liable to affect the continuity of the public service.

Most comparable legal systems organise the same thing. In Morocco, Law no. 54-05 on the delegated management of public services provides that return assets must revert compulsorily to the delegating authority at the end of the delegated management (article 16), that the annexes to the contract include in particular an inventory of the movable and immovable assets made available to the delegated operator (article 12), and that the delegating authority may request communication of, or inspect, any document held by the delegated operator relating to the performance of the operations of the delegated management (article 17).

On paper, then, the public owner is very well provided for. It owns without having paid, it owns without having stipulated, and it recovers at expiry. The whole question is what exactly it recovers.

What a good contract guarantees in addition: condition

Ownership says nothing about condition. A network can return to the delegating authority in a state that makes it unusable, and the right of ownership changes nothing. That is why serious concession contracts add an obligation that the law does not impose: an obligation of condition at handback.

The concession of the British high-speed line HS1 offers a public and readable example of this, which is rare. Its holder, in its own words, holds "the concession from the UK government to operate, maintain, renew and replace the HS1 assets until 31 December 2040", under the supervision of the rail regulator. Its obligations are published, and it reports on them in documents open to all. On the stations, it writes:

Under the Station Leases [...], we have a number of asset stewardship obligations, including keeping the stations in "good and substantial repair and condition" at all times during the concession, including on handback to the government at the end of the concession.

On the funding of renewals, the mechanism is even more remarkable. The holder pays into an escrow account, and, under the regulator's interpretation of the annuity framework set out in the concession agreement, it has been required in previous periodic reviews to:

Fully fund renewals over a 40-year period; Ensure we renew assets in accordance with best practice to ensure an equivalent handback of assets at the end of the concession; and Ensure the escrow account reaches zero balance at the end of the 40-year period with no negative balances during this period.

(The 2025 statement relaxes the last condition slightly for the current control period, allowing small negative balances in later years of the modelling.)

Here is a contract that does far better than the law. It does not merely return the works: it requires them to be returned in an equivalent condition, it funds this obligation through an escrowed and verifiable mechanism, and it places the whole under the control of a regulator. This is, to my knowledge, roughly the state of the art.

Yet let us reread what is promised. "An equivalent handback of assets". The stations in good and substantial repair and condition. Both commitments concern things, and the condition of those things. Neither concerns knowing how to use them.

What neither the law nor the contract returns

A structure is not operable because it is in good condition. It is operable because someone knows what to do with it.

Operating a treatment plant, a heat network, a railway line or a desalination plant presupposes a set of things that appear on no inventory. Knowing which settings hold and which drift. Knowing which piece of equipment fails in spring and why. Knowing which part has only one supplier, and which one. Knowing which written procedure no longer matches reality on the ground since a modification made six years ago. Knowing whom to call at three in the morning. In asset management, this set has a name: it is operational capability. It has considerable value, and it has no designated owner.

It is not a return asset, since it is not an asset. It is not a deliverable, since no contract orders it. It is not depreciable, since it is not on the balance sheet. And it cannot be transferred by decision, since it resides in people and in collective habits.

The practical consequence can be verified at the end of every contract. The day the outgoing operator leaves, the works remain and the capability departs. The delegating authority finds itself the owner of assets it does not know how to operate, and it has only two options: renew the outgoing operator, which considerably weakens its negotiating position, or pay an incoming one to rebuild what already existed. In both cases it pays twice for the same thing: once during the contract, and a second time to recover its effect.

This is not a manoeuvre by the delegated operator. It is a void. No one has ever written that operational capability must come back, because no one has ever treated it as something that could come back.

The blind spot of the reference framework

One might think that the profession has dealt with the question elsewhere. It has not, and this can be checked.

The Asset Management Landscape of GFMAM, the Global Forum on Maintenance and Asset Management, is the reference map of the discipline. Its third edition, published in 2024, describes forty subjects in seven groups. It made an interesting move: the subject devoted to the supply chain, which sat in the "Organisation & People" group of the second edition, has moved to the "Delivery" group, a sign that subcontracting is now understood as execution of the life cycle rather than as a support function.

But let us read what this subject covers, in the Landscape's own terms:

Supply Chain Management is the process used by an organization to ensure the provisioning of all equipment, tools, and resources to perform Asset Management activities are aligned with the Asset Management objectives. [...] Supply chain management includes all the activities necessary to acquire, manage, maintain, and enforce contract and supplier management over the entire life cycle of an asset, taking total cost of ownership into consideration. It includes purchasing, inventory, warehousing, and contracting (authoring, negotiations, adoption, definition of requirements, appraisal and selection of contractors, outsourcing-insourcing strategies, and claim management). [...]

It is a supply subject. It deals with buying, storing, contracting and monitoring a supplier. "Outsourcing-insourcing strategies" appear in it, which is the closest mention of our subject, but as an organisational choice, not as a question of residual capability. The case where the contractor operates the asset and the owner does not know how to do so is addressed by none of the forty subjects.

ISO 55001:2024 is no more explicit, and this is even more surprising since it devotes a whole clause to external providers. Its clause 8.3 asks the organisation to ensure that what is provided by external providers is controlled, and to determine three things: the processes, technologies, products, services and activities that are outsourced, including their scope, boundaries and interfaces; the responsibilities and authorities within the organisation for managing what is outsourced; and the processes for, and the scope of, sharing knowledge and information between the organisation and its service provider or providers (paraphrased from the French edition of the standard).

The third point is remarkable, and little commented on. The standard requires the organisation to define the scope of knowledge sharing with its provider. This is exactly the right place to deal with our question. But the requirement concerns the definition of a scope, not its content: an organisation that decided, in full awareness, that nothing would be shared would satisfy the letter of points a) to c). The clause does refer, for what is outsourced, to the requirements of clauses 7 and 8, including the retention of existing knowledge, but it does not say what must come back to the owner at the end of the contract. The standard asks for an answer. It does not say which one.

As for ISO 37500, the only international standard devoted to outsourcing, its official abstract states that it is intended "to enable organizations to enter into, and continue to sustain, successful outsourcing arrangements throughout the contractual period". Throughout the contractual period. The moment that concerns us, when the contract ends and one has to carry on without the provider, falls just after the sentence.

I have read only its official abstract and scope, not the full text, and I say so rather than conclude too quickly. But the announced scope is clear.

Asset data, a particular case of a general problem

The previous issue of this collection dealt with asset data, and with the fact that the exit right of the European Data Act organises the return of tables without organising the return of the model. What we see here is the same phenomenon, one level up.

Data is the most tangible part of operational capability, and therefore the easiest to claim. It is also the only part whose return is beginning to be organised, by the European regulation on one side and, in public service delegations, by the information obligations that the laws require the contract to provide for. But it is only one part. A delegating authority that recovered all of its operator's data, in a perfect format, would still not know how to operate its installation. It would know what happened. It would not know what to do.

The hierarchy therefore has three levels, and each is less well protected than the one before. The works come back; the law sees to it. Their condition comes back, if the contract is well written. The data is beginning to come back, if one thinks of requiring it. The capability does not come back, and nothing provides for it to come back.

ISO 55013, published in July 2024 and devoted to the management of data assets, illustrates this ceiling well. It finally recognises asset data as an asset in its own right. And its official abstract immediately specifies that it "does not provide methodologies to derive or appraise value for data assets", nor to derive financial values for them. The asset is named, and its valuation is given up. An asset without a recorded value cannot be defended in an end-of-contract negotiation.

What those who succeed do: the intelligent client

If the problem were insoluble, there would be nothing to write. Yet some delegating authorities and some concessionaires solve it, and they solve it in the same way: by deliberately funding a function that no one asks of them.

The HS1 concession holder outsources almost all of its operations. It does not hide this; it claims it, and at the same time describes what it has chosen not to outsource:

We operate through an outsourced model, in which we lead the supply chain as an intelligent client. We have a good understanding of our asset and our requirements, including our long-term asset stewardship obligations, and we challenge our suppliers to improve their practices and deliver efficiently.

And, more explicitly still:

In order to work effectively with our suppliers, we act as an intelligent client and have developed our in-house capability to engage, oversee, direct and challenge our supply chain to deliver more efficiently and effectively.

The verb that matters is the last one. Challenge. A client able to challenge is a client that knows what it should obtain, and therefore a client that has kept enough technical competence to judge. This is exactly what delegation destroys when it is not thought through, and what this one has chosen to preserve.

The same organisation adds, about its safety obligations, a sentence that should appear in every delegation contract:

This does not absolve us of our duties and we will continue to demonstrate that we are acting as an informed and intelligent client [...]. As the client organisation, we remain focused on longer-term safety threats and opportunities through capability, maturity and innovation whilst the supply chain manages the shorter-term risk.

Here is a clear division of roles, and it is the reverse of the one usually observed. The provider takes the short term, because it is equipped for it and its contract has a horizon. The owner keeps the long term, because it is the one that will still be there. In most poorly designed delegations, exactly the opposite happens: the delegated operator builds the long view, without which it could not recover its investment, and the delegating authority ends up arbitrating emergencies it only half understands.

A caveat of honesty is needed here. These sentences are written by the concessionaire itself, in a document addressed to its regulator. They are statements, not independent findings. They count as a statement of doctrine, and because the regulator receives them and comments on them publicly, not as proof of results.

The forty-year test

The second mechanism deserves to be known well beyond British railways, because it elegantly solves the most classic problem of the end of a contract.

A delegated operator whose contract ends in five years has no economic interest in committing an expense whose benefit will appear in ten. It even has the opposite interest. This is the phenomenon regulators call end-of-concession skimming, and it is almost impossible to sanction, because it shows itself only through things that were not done.

The general duty that the HS1 concession agreement imposes on its holder gets round the obstacle. The holder must secure the operation, maintenance, renewal and replacement of the infrastructure in accordance with best practice, in a timely, efficient and economical manner, and, in the words of its own asset management statement:

as if we were responsible for the stewardship of the HS1 railway infrastructure for 40 years following the date that any such activities are planned or carried out.

The clause does not ask the holder to behave well. It imposes a legal fiction on it: to act as if it remained responsible forty years after each of its acts. A decision taken in the last year of the concession must therefore be defensible forty years later, before someone else. The decision-maker's horizon is detached from the horizon of its contract.

It is a transferable device, and a cheap one. It fits in one sentence, costs nothing to stipulate, and shifts the burden of proof: it is no longer for the delegating authority to show that an expense should have been incurred, but for the delegated operator to show that its choice would stand the test of time.

What a delegating authority can demand

Nothing that follows presupposes any particular balance of power, nor any change in legislation. These are contract terms; they are negotiated at signature, and they cost all the less the earlier they are asked for.

• Name operational capability as an object of the contract. As long as it is not named, it does not exist legally. A definition is enough: the body of knowledge, procedures, settings, histories and supplier relationships without which the work handed back cannot be operated under the conditions of the service.
• Require a capability inventory, distinct from the inventory of assets. The inventory of works is already compulsory in most regimes. It says nothing about what one must know to make them run. A second document, kept up to date, records the procedures actually applied, the critical settings and their justifications, the skills required and the single-source supplier dependencies.
• Provide for the overlap period before the end, not after. A reversibility that starts on the last day is no reversibility. Twelve to eighteen months before expiry, the outgoing operator trains the delegating authority or its future provider, in real conditions and under supervision. This period is stipulated, dated and paid for in the original contract, when the outgoing operator does not yet know whether it will be renewed and therefore has every interest in doing it well.
• Test the takeover during the contract. A takeover capability that has never been tried is not a capability; it is a clause. A periodic exercise, every two or three years, during which the delegating authority's teams carry out a routine operation under supervision, turns a promise into a verified fact, and reveals dependencies at the moment when they can still be corrected.
• Fund the intelligent client function, and protect it in the budget. This is the hardest measure, because it is the only one that visibly costs money. It consists in maintaining within the delegating authority a technical competence that produces nothing directly, whose sole function is to understand, supervise and challenge. It is also the only one that makes all the others enforceable: with no one to read the capability inventory, it is just a binder.
• Stipulate the long-term test. Take up the clause of the British concession, adapting the duration to the nature of the asset: the delegated operator acts as if it remained responsible for asset stewardship for a set period following each of its acts.

A remark on the order. The fifth measure conditions the others, but it is the one that budget decisions remove first, because it has no measurable short-term result. That is exactly why it must appear in the delegation contract itself, funded as a line of the delegation, and not in the delegating authority's operating budget, where it will be weighed every year against more visible expenses.

What we do not know, and what matters

Three uncertainties, and keeping quiet about them would weaken what precedes.

The cost of rebuilding is not quantified. The argument that the same capability is paid for twice is sound in reasoning and consistent with what practitioners observe, but I have found no public study that measures it. As long as a figure does not exist, this study describes a mechanism; it does not quantify a loss. It is a limitation, and it is also a research programme.

The legal characterisation of capability remains open. Can an operational capability be attached to the regime of return assets by a contract term, given that these regimes admit movable assets that contribute substantially to the operation of the service? The argument exists, it turns on importance rather than nature, but it has not been settled to my knowledge, and it is a matter for the opinion of a specialised lawyer in each legal order concerned.

The British case is a declarative source. The commitments quoted are written by the concessionaire in a document addressed to its regulator. They describe a doctrine publicly assumed and controlled by a third party, which gives them weight, but they do not constitute an independent evaluation of results.

What holds remains. The law returns the walls, and it returns them well. The contract, when it is good, guarantees that they will be in good condition. Data is beginning to come back, for those who think of requiring it. But knowing how to use them has no owner, no inventory, no recorded value and no obligation of return. This is not a technical fatality and it is not a fault of the delegated operators: it is a contractual void, and it can be filled by contract terms that already exist elsewhere. The question is not whether the work will come back. It is who, on that day, will know how to make it run.

Sources and references

All sources were consulted and verified between September and October 2026. The [n] cross-references in the text point to this list.

1. Conseil d'État, Assembly, 21 December 2012, Commune de Douai, no. 342788, published in the Recueil Lebon. Grounds 2 to 4. Légifrance, CETATEXT000026810748. https://www.legifrance.gouv.fr/ceta/id/CETATEXT000026810748/
2. HS1 Ltd, Five Year Asset Management Statement for Control Period 4, response to the Final Determination, 3 February 2025. Sections 1.2.1, 1.2.2 "Regulation", 1.2.3 "Supply chain", 8.1, 10.1 and 13.6.1. Public document submitted to the Office of Rail and Road. https://stpancras-highspeed.com/wp-content/uploads/2025/02/HS1-Ltd-5YAMS-February-2025-FD-Response-1.pdf
3. HS1 Ltd, Asset Management Assurance Statement 2023-24, annual statement of compliance with ISO 55001 submitted to the regulator.
4. Concession Agreement between the UK government and HS1 Ltd, schedule 10 on asset stewardship obligations and periodic reviews, cited in the two preceding documents. Concession running until 31 December 2040.
5. HS1 station leases, schedule 10 of the HS1 Lease and the Ashford Deed, obligations to keep the stations "in good and substantial repair and condition" including on handback, cited in the same document.
6. ISO 55001:2024, Asset management, Asset management system, Requirements, second edition, July 2024, clause 8.3 "Externally provided processes, products, technologies and services". https://www.iso.org/standard/83054.html
7. ISO 37500:2014, Guidance on outsourcing. Official abstract and scope. https://www.iso.org/standard/56269.html
8. ISO 55013:2024, Asset management, Guidance on the management of data assets, first edition, July 2024, ISO/TC 251. https://www.iso.org/standard/82455.html
9. GFMAM, Asset Management Landscape, third edition, 2024, subject 6.8 "Supply Chain Management", group Delivery. In the second edition, the corresponding subject was no. 26, "Procurement & Supply Chain Management", in the Organisation & People group.
10. Regulation (EU) 2023/2854 of 13 December 2023 on harmonised rules on fair access to and use of data (Data Act), chapter VI. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32023R2854
11. Moroccan Law no. 54-05 on the delegated management of public services, promulgated by Dahir no. 1-06-15 of 15 Moharram 1427 (14 February 2006), Bulletin officiel no. 5404 of 16 March 2006, official French translation edition, pp. 506-510, articles 12, 16 and 17.
12. Mqam, N. Y. (2026). Asset data: exit becomes free, the model stays captive. Cahiers de technopolitique, no. 3, English version. https://doi.org/10.5281/zenodo.23207208 (French original: https://doi.org/10.5281/zenodo.23207207).

Download

Read the document (PDF)

8 pages · English · 12 sources · 100 KB
Fully selectable and citable text
The document carries its own DOI, reserved before export
CC BY-NC-ND 4.0 licence, no registration

Bibliographic record

Title

Delegated operation: the works come back, the capability does not

Author

Nizar Younes Mqam, ORCID 0009-0008-5549-7620

Serial publication

Cahiers de technopolitique, no. 4, Study format

ISSN

3131-7979

Publisher

Nizar Younes Mqam, Tourcoing, France

Publication date

19 September 2026

Revision

7 October 2026: every quotation, figure and attribution checked against its primary source; dates, figures, attributions and references corrected, cuts in quotations marked [...]. The argument is unchanged. Version 1.2, same day: Moroccan provisions stated from the official text of Law no. 54-05 (articles 12, 16 and 17), scope of the delegating authority's right of communication specified, a drafting note that had become moot removed. Version 1.3, 10 October 2026: the number of uncertainties announced in the last section was corrected after that note was removed.

Language

English

DOI

10.5281/zenodo.23284423
All versions : 10.5281/zenodo.23199880

Permanent repository

Zenodo, hosted by CERN

Licence

CC BY-NC-ND 4.0

Other languages

Français (original)

To cite this study

Mqam, N. Y. (2026). Delegated operation: the works come back, the capability does not. Cahiers de technopolitique, n° 4. https://doi.org/10.5281/zenodo.23284423

Read also

No. 3 · Study · September 2026

Asset data: exit becomes free, the model stays captive

The previous issue: the EU Data Act returns the tables at exit, not the model of the asset register.

14 pages · 28 sources · PDF 125 KB
DOI 10.5281/zenodo.23207208

Also inFrançais

Read the record and download

No. 5 · Study · September 2026

The cost of taking back: you cannot price what you have stopped seeing

The cost of taking back: half of the public authorities surveyed keep no register of the assets under their contract, so the cost of taking an operation back in-house cannot be measured.

7 pages · 8 sources · PDF 97 KB
DOI 10.5281/zenodo.23207038

Also inFrançais

Read the record and download

No. 7 · Study · October 2026

ISO 55001 and the AI agent: who signs before the auditor?

When the last operators who ran the column by hand have gone, who will know how to take back the valves? The capability that does not come back, AI agent version.

11 pages · 30 sources · PDF 115 KB
DOI 10.5281/zenodo.23207212

Also inFrançaisالعربية

Read the record and download