Study no. 5 · Sovereignty and asset data

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

Nizar Younes Mqam ORCID 0009-0008-5549-7620

Study · 19 September 2026 · Cahiers de technopolitique, no. 5

The previous issue argued that a delegating authority pays twice for the same operational capability, and acknowledged that it had found no public study measuring it. This issue went looking for the figure. It does not exist, and the reason is more interesting than the figure itself: the British public audit institution found, through a survey covering one hundred and seven contracts, that half of the public authorities that responded keep no register of the assets of their contract. You cannot invoice what you have stopped seeing.

The question left open

A study that leaves a question open must come back for it. Issue 4 of this collection argued that the operational capability of a delegated asset has no owner, no inventory and no obligation of return, and that the delegating authority therefore pays for it twice: once during the contract, and a second time to rebuild it afterwards. It added, in its section on uncertainties, that the argument was sound in reasoning but that I had found no public study that measures it.

I went looking for that measurement. I did not find it, and I will explain why I now believe it does not exist, rather than having simply escaped my searches.

Along the way, I found something better. Not the cost itself, but the state of what would make it possible to calculate it, measured by an institution whose job this is, on a sample that is not open to dispute.

The scale, as a national auditor measures it

On 5 June 2020, the National Audit Office, the British public audit institution, published a report to Parliament entitled Managing PFI assets and services as contracts end. It concerns public-private partnership contracts, known in the United Kingdom as the private finance initiative, at the moment when they end and the assets return to the public authority. This is the subject of the previous issue, examined this time by an auditor with powers of investigation.

The scale first, because it explains why the subject is not anecdotal:

• More than 700 partnership contracts are in operation in the United Kingdom, with a capital value of 57 billion pounds. The study covers the 571 English contracts.
• The payments still due under these contracts amount to 168 billion pounds over thirty years from 2020-21, or 5.6 billion a year.
• 204 contracts expire within ten years, covering assets with a capital value of 11.7 billion pounds.
• The pace accelerates sharply: an average of 10 expiries a year up to 2024-25, then 31 a year between 2025-26 and 2029-30.
• Over seven years, 72 contracts end in England, and 3.9 billion pounds of assets return to public ownership.

So this is not a problem of a distant end of cycle. It is a wave that has begun, and that triples in intensity as I write.

Half of them have no inventory

Next comes what the auditor measured among the public authorities themselves, by questionnaire. I quote the findings in its own terms, because they speak for themselves:

30% of survey respondents are not monitoring annual maintenance spending and 50% do not maintain an asset register.

Half of the public authorities that responded keep no register of the assets of their contract. This sentence must be read twice. These are not assets whose existence they are unaware of: they are, for the most part, assets that will come back to them at expiry, and that they do not record. Under French law, as the previous issue recalled, they would even belong to them from the moment of construction.

The auditor continues, and what follows partly explains the previous finding:

Many PFI contracts, particularly those signed before 2000, contain contractual limitations over what data can be requested from the SPV, such as maintenance expenditure and the ongoing assets' condition. Around 35% of respondents stated they had insufficient access rights to monitor the maintenance programme adequately.

And, more starkly still:

There is also evidence that PFI investors and sub-contractors are not cooperating with authorities to provide information – 20% of survey respondents who had asked for information considered the contractor had been uncooperative, with data requests being ignored or denied on grounds of not being a contractual obligation.

"On grounds of not being a contractual obligation." Issue 3 of this collection argued that the right of exit is defined by subtracting what the provider claims, and issue 4 that operational capability is not a deliverable because no contract orders it. Here is the same mechanism, observed by an auditor, in the mouths of the contractors themselves, and set against public authorities to which these assets are due to return.

One last figure completes the picture: about 45% of the contracts in the sample were signed before HM Treasury introduced standardised clauses, and therefore carry, the auditor says, a higher risk of ambiguity that may lead to disputes. Nearly one contract in two was drafted before anyone knew what to write in it.

Why the figure does not exist

We can now answer the question left open, and the answer is not that no one thought of doing the calculation.

Calculating the cost of rebuilding an operational capability requires knowing three things: what one had before, what one has after, and what was spent in between. Yet the auditor has just established that the first is unknown in one case out of two, that the third is not monitored in one case out of three, and that the information needed for both is sometimes contractually out of reach.

The cost is therefore not invisible because it is small. It is invisible because the instrument that would make it possible to measure it is precisely the thing that has been lost. An owner without an inventory cannot observe an inventory gap. An owner without a maintenance history cannot establish that it paid twice for the same intervention. The loss hides in the very tool that would reveal it.

This is a familiar figure in public accounting: a cost that finds no budget line appears nowhere, even when it is paid. It dissolves into an operating budget, an assistance contract, a study contract, the extra cost of a tender received. Each of these items is visible separately; the whole they form is not.

I must be precise about what I am claiming here. I am not saying that the figure is impossible to produce. I am saying that it cannot be produced after the fact, by analysing records, in organisations that did not keep those records. It can only be produced prospectively, by instrumenting the contract expiries to come. And, according to the auditor's projections, there will be thirty-one of them a year on average between 2025-26 and 2029-30, in a single country.

What can be measured all the same: the symptoms

Unable to measure the illness, the auditor measured its signs. They converge.

The gap between what is expected and what is found. Just over 70% of respondents whose assets return to the public authority say they expect to receive the assets in the condition required by the contract, or better. But among the nine authorities whose contract has already ended and which have taken the assets back, four out of nine were not satisfied with the condition in which they received them.

The two figures concern different populations and the second is small, as I point out. But the gap between an expectation of 70% satisfaction and an experience of 44% dissatisfaction is exactly what one would expect from a system where information is missing: one is confident as long as one does not look.

The nature of the disputes expected. More than one third of respondents consider a dispute likely as expiry approaches, and 86% of these disputes are expected to concern the quantity of rectification work, ahead of its cost (75%) and the question of who should pay for it (46%). The first ground is neither the price nor the payer: it is how much there is. It is first of all an inventory quarrel, and it only happens for lack of a shared inventory.

The shift onto other budgets. The auditor gives a concrete order of magnitude: one authority reports a maintenance backlog of around 90 million pounds on its portfolio of non-PFI schools, and taking over former PFI schools may add to it. The cost does not disappear; it changes line.

Skills. On the conduct of contract expiry itself, the auditor notes that about 30% of respondents expect not to have enough staff, that about 25% consider they lack the necessary in-house skills, and that 60% plan to hire consultants. It adds a remark that holds beyond its subject: the skills required to deliver an exit, contract negotiation and asset management, differ from those needed to manage the day-to-day running of the contract. One does not become able to leave by having been able to stay.

Finally, it notes that the piecemeal approach to hiring consultants "may not represent value for money in the long term". It is a cautious way of saying that one is buying back, firm by firm, a competence one once had.

The perverse incentive, named by the auditor

It remains to understand why the condition of the assets deteriorates precisely as expiry approaches. The auditor describes the financial mechanism, and it is instructive because it accuses no one.

In more than half of the contracts in the survey, the project company builds up a fund dedicated to lifecycle replacement, fed by a share of each charge paid by the authority. It bears the risk that this fund is insufficient. But, the auditor writes, "any surpluses are typically retained by the SPV and paid out to investors".

This construction has a virtue: it encourages good maintenance, since a well-kept estate requires fewer unplanned replacements and therefore leaves a larger surplus. And it has a symmetrical flaw, which the auditor names without hesitation:

Equally, there is a perverse incentive to underinvest in assets – some authorities raised concerns that SPVs are 'sweating' the assets and making them last longer than originally planned.

The verb used, sweating, is the trade's own. It means drawing from a piece of equipment the last year of service it can give, pushing its renewal beyond the end of the contract. It is perfectly rational for whoever is leaving, and perfectly costly for whoever stays.

The previous issue described a device that neutralises exactly this incentive: the clause of the British rail concession that requires its holder to act as if it remained responsible for asset stewardship for forty years after each of its acts. In the light of this report, one sees better what it costs not to have that clause.

The only existing costing, and its method

There is nevertheless a costing of what information that does not flow costs. It is old, it is American, and it does not concern contract expiries. But its method points the way.

In 2004, the United States National Institute of Standards and Technology published a study, carried out for it by RTI International and the Logistics Management Institute, on the cost of inadequate interoperability in the capital facilities industry. It puts it at 15.8 billion dollars a year for 2002, of which two thirds are borne by owners and operators, and 85% of their share falls in the operation and maintenance phase.

What interests us here is not the amount but the way it was obtained. The authors do not look for an expense identifiable in accounts, because they know it is not there. They build a counterfactual scenario: a world in which data exchange would be seamless, which implies, they write, that "information need only be entered into electronic systems only once". Then they measure the gap between that world and ours, item by item, through interviews and a survey, distinguishing avoidance costs, mitigation costs and delay costs. Manual re-entry and data translation items appear explicitly in their tables.

Applied to our question, the method would give this: not "how much did the takeover cost", a question with no answer, but "how much would the same takeover have cost if the inventory, the specification and the history had been kept and handed over". The gap is the cost sought.

A caveat, which I state myself. The NIST figure is more than twenty years old, it is American, and it concerns buildings and industrial facilities, not networks or delegated services. I do not transpose it, and I do not update it: that would be manufacturing an amount that no one measured. I keep its method, not its result.

What it would take to put a figure on it

A measurement protocol would fit in a few points, and the wave of expiries ahead offers the ground for observation. What follows is not a proven method; it is a proposal, and I offer it as such.

• Set the baseline at the start, not at the end. A record of inventory, specification and history, drawn up at signature, is the only reference against which a gap can be observed. Without a baseline there is no measurement, only impressions.
• Isolate a "takeover" budget line. As long as assistance fees, upgrade studies, consultancy services and tender surcharges remain scattered, the total never appears. A single line, opened at signature and fed until two years after the end, would make the amount simply readable.
• Distinguish three kinds of expenditure, in the manner of NIST: what is spent to avoid the problem, what is spent to repair it once it has occurred, and what the delay it causes costs. The three are not managed in the same way.
• Measure on a cohort, not on a case. A single contract teaches nothing, because each expiry has its own circumstances. Thirty-one expiries a year in a single country form a cohort large enough to tell chance from structure.
• Publish the protocol before the results. This is what separates a measurement from an argument. A method announced in advance, applicable by others, produces a figure that can be challenged, and is therefore useful.

None of these five steps requires reform. The first four are a management decision, the fifth a publication decision. What is missing is not the power to measure but the interest in doing so: the cost is borne by the successor, and whoever would inflict it is no longer there to answer for it.

What we do not know, and what matters

Four caveats, and they are important because this study concludes on an absence.

An absence of evidence is not evidence of absence. I looked for a study putting a figure on the cost of rebuilding an operational capability and did not find one. The reasoning developed here explains why it would be hard to produce, which makes its absence plausible, but does not prove it. If a reader knows of one, it interests me more than anything else in this document.

The report relied on covers a single country and a single legal instrument. British partnership contracts have their own features, notably the project company and its lifecycle fund, which have no exact equivalent elsewhere. The findings on inventory and access rights seem to me transferable, because they stem from the structure of delegation and not its legal form, but that is a judgement, not an established result.

The percentages quoted rest on unequal numbers. The survey covers 107 contracts, and some findings concern much smaller subsets: the "four out of nine" on satisfaction at handback is the clearest example. I quoted it because it is the only feedback available on contracts that have actually ended, not because it would be statistically robust.

The proposed protocol has not been tested. It is consistent with the NIST method, it is feasible, and it has not been applied anywhere to my knowledge. It stands as a working proposal, not as a validated method.

What remains, once these caveats are stated, fits in one sentence. The cost of taking back an operation exists, it is paid, and it is written down nowhere, because the organisation that bears it stopped, by delegating, keeping the books that would make it visible. The question is therefore not how much it costs. It is why no one wants to know, and the answer is simple: whoever would pay to find out is never the one who would benefit from it.

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. National Audit Office, Managing PFI assets and services as contracts end, report by the Comptroller and Auditor General, HC 369, session 2019-21, ordered by the House of Commons to be printed on 3 June 2020, published on 5 June 2020, 62 pages. Key facts, summary, parts one to three, and appendix four on survey responses. Cabinet Office and HM Treasury. https://www.nao.org.uk/reports/managing-pfi-assets-and-services-as-contracts-end/
2. Infrastructure and Projects Authority, Expiry Toolkit: Further Guidance and Resources, version 1.0, February 2022. Collection of resources published following the previous report. https://assets.publishing.service.gov.uk/media/621c88228fa8f5490d52eed2/Expiry_Toolkit_-_Further_Guidance_and_Resources.pdf
3. NIST GCR 04-867, Cost Analysis of Inadequate Interoperability in the U.S. Capital Facilities Industry, August 2004, prepared for the National Institute of Standards and Technology by RTI International and the Logistics Management Institute, 194 pages. Executive summary, sections E.1 and E.2, and section 6.5. https://nvlpubs.nist.gov/nistpubs/gcr/2004/nist.gcr.04-867.pdf
4. Conseil d'État, Assembly, 21 December 2012, Commune de Douai, no. 342788, published in the Recueil Lebon, ground 2. https://www.legifrance.gouv.fr/ceta/id/CETATEXT000026810748/
5. ISO 55001:2024, Asset management, Asset management system, Requirements, second edition, July 2024, clauses 7.6 and 8.3. https://www.iso.org/standard/83054.html
6. HS1 Ltd, Five Year Asset Management Statement for Control Period 4, February 2025, public document submitted to the Office of Rail and Road. General asset stewardship duty with a forty-year horizon. https://stpancras-highspeed.com/wp-content/uploads/2025/02/HS1-Ltd-5YAMS-February-2025-FD-Response-1.pdf
7. Mqam, N. Y. (2026). Delegated operation: the works come back, the capability does not. Cahiers de technopolitique, no. 4, English version. https://doi.org/10.5281/zenodo.23207210
8. 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

Download

Read the document (PDF)

7 pages · English · 8 sources · 97 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

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

Author

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

Serial publication

Cahiers de technopolitique, no. 5, Study format

ISSN

3131-7979

Publisher

Nizar Younes Mqam, Tourcoing, France

Publication date

19 September 2026

Language

English

DOI

10.5281/zenodo.23207038
All versions : 10.5281/zenodo.23207037

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). The cost of taking back: you cannot price what you have stopped seeing. Cahiers de technopolitique, n° 5. https://doi.org/10.5281/zenodo.23207038

Read also

No. 4 · Study · September 2026

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

The previous issue: the law returns the works and the contract their condition; nothing returns the capability to operate them.

8 pages · 12 sources · PDF 100 KB
DOI 10.5281/zenodo.23284423

Also inFrançais

Read the record and download

No. 3 · Study · September 2026

Asset data: exit becomes free, the model stays captive

The Data Act makes switching provider free from 2027; what you can take with you is another matter.

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

Also inFrançais

Read the record and download

No. 7 · Study · October 2026

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

When an AI agent made the decision on a critical asset, what do you show the auditor? Nine real cases, one standard reread, five items to demand.

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

Also inFrançaisالعربية

Read the record and download