Measure 3.09 — Reinternalising AFIT France without confusing policy flows with savings

A structural reform, not the disappearance of €3.75 billion

Measure 3.09 proposes to bring the Agence de financement des infrastructures de transport de France — AFIT France or AFITF — back inside the central administration of the transport ministry. It is an ideal pilot because it forces the reform plan to separate two figures that are often confused in debates about public agencies: the amount of public money flowing through an agency and the cost of keeping the agency itself as a separate institution. For 2026, earmarked revenue and budget credits associated with AFIT France are estimated at roughly €3.75 billion. That money finances transport infrastructure. If the agency disappears as a legal entity, the railways, roads, public transport projects and contractual commitments do not disappear with it. A ministry that takes over the mission must also take over the resources required to finance that policy.

The real question is organisational. Does France need a separate legal entity, board, accounting framework and authorising officer to distribute funds whose strategic direction is already set by the State? The French Senate’s inquiry into “agencification” answered that question quite directly for several small financing agencies. It recommended that AFITF’s activities be reinternalised within the central administration of the Ministry of Transport. That recommendation does not prove that the change will generate a large saving, but it makes the reform a concrete case rather than a theoretical attack on “agencies” in the abstract.

What AFIT France is in law

The French Transport Code defines AFITF as a national administrative public establishment with legal personality and financial autonomy, supervised by the minister responsible for transport. Its purpose is to contribute to the financing of national and territorial infrastructure: road, rail, inland waterways, ports, collective transport and certain partnership projects. It can award investment grants, repayable advances and contributions to funds. It has a board and receives budget resources and earmarked revenue under the relevant financial legislation.

That autonomy serves a function. A dedicated financing vehicle can make transport investment more visible, support multi-year commitments and protect some long-term projects from short-term reallocations. Defenders of AFITF are therefore not defending an empty shell. The reform must answer a more precise question: does that useful programming function require a separate public establishment, or can the same guarantees be reproduced inside the State budget with a clear programme, multi-year trajectory and parliamentary reporting?

A very small structure compared with the flows it handles

Evidence gathered by the Senate inquiry is striking. AFIT France was described as having four permanent staff while relying heavily on the central administration of the transport ministry. The Senate also highlighted the difficulty of reading the agency’s full cost. A very small personnel line in the agency’s own accounts can understate the true administrative resources used if staff and expertise are carried by the parent ministry. That is precisely why the reform must use full-cost accounting: if the ministry already provides much of the administrative work, abolishing the agency does not magically create a saving equal to those ministry resources.

At the same time, this arrangement raises a legitimate governance question. When a body with a handful of permanent staff manages several billion euros of flows, backed by a board and separate accounts, the State should be able to explain the specific value created by that institutional separation. If files are largely prepared within the transport administration, if the ministry provides expertise and if policy priorities ultimately come from the Government, reinternalisation can shorten the chain of accountability: the ministry programmes, authorises, reports and answers directly to Parliament.

What reinternalisation would actually require

The reform cannot consist of closing the agency on Friday and telling the ministry to improvise on Monday. Existing funding agreements, multi-year commitments, earmarked taxes, payment schedules, contracts, archives, accounting data and authorising responsibilities need an orderly transfer. Provisions in the Transport Code establishing and governing the public establishment would need to be amended or repealed. Finance legislation would have to redirect revenue currently earmarked to the establishment towards a State programme or other mechanism that preserves continuity of infrastructure funding.

A multi-year programming mechanism is essential. Without it, reinternalisation might improve administrative simplicity but weaken investment visibility. One option would be a clearly identified budget programme accompanied by a multi-year trajectory voted or debated by Parliament and a detailed annual implementation report. The objective would be to preserve the useful part of AFITF — long-term visibility — while removing a separate institutional layer if that layer does not demonstrate enough additional value.

Who decides, pays, implements and audits?

Who decides? The Government initiates the reinternalisation and regulatory changes; Parliament intervenes where legislation and earmarked revenue need to be changed. Who pays? Public resources for infrastructure remain public resources and continue to fund projects unless policy itself changes. Who implements? The transport ministry and its competent directorates take over programming, accounting and agreement management. Who audits? Parliament controls appropriations and earmarked revenue, the Cour des comptes can review execution, and the ministry should publish commitments, payments and performance indicators.

Reproducible costing: measure the institution, not the infrastructure budget

The baseline must be the full cost of the agency’s own institutional layer. That means identifying remuneration borne directly or indirectly for the agency’s functioning, board and governance costs, accounting, audit, dedicated information systems, premises, external services, banking arrangements if any, and administrative time that exists specifically because a separate establishment exists. The calculation must then subtract the costs the ministry would incur after taking over the function: staff, accounting control, reporting, systems and agreement management.

There may also be non-cash benefits: fewer internal agreements, shorter processing times, clearer responsibility and less duplicated reporting. Those benefits should be measured separately rather than converted into euros without evidence. The first year may even produce a negative fiscal effect because systems, contracts and records need to be migrated. Delta-Sierra should therefore show three distinct lines: recurring structural saving, one-off transition cost and measured administrative performance gain.

What is known and what still needs evidence

Known: AFITF’s legal status and mission, the approximate €3.75 billion 2026 resource/credit envelope, and the Senate inquiry’s recommendation to reinternalise it. Well documented: the extremely small permanent staff and the reliance on central administration. Still to be documented precisely: the annual full cost that would actually disappear if the separate legal entity were removed. Until that figure is produced from accounting and staff data, any precise claim of tens or hundreds of millions of euros of annual savings would be speculative.

Risks, objections and safeguards

The strongest objection is loss of ring-fencing. A dedicated agency can make infrastructure resources visible and support multi-year planning. The safeguard should therefore be budgetary: a named programme, multi-year trajectory, transparent list of projects and annual reporting. A second objection is that central administration could be slower. The pilot should answer that with measurable processing deadlines and publication of the time from funding decision to contractual commitment and payment. A third objection is that all four staff members and the same tasks may simply move to the ministry. That is entirely possible; if so, the cash saving may be small and the site should say so.

The measure can still be valuable if accountability improves. Today, a financing flow may be prepared by the ministry, formally handled through an agency structure and funded with earmarked resources. After reinternalisation, the minister and administration would be more directly accountable for the programme before Parliament. State reform should therefore be judged not only by euros saved but also by the number of institutional interfaces removed, the clarity of responsibility and the speed and traceability of decisions.