Highways : $38 in dividends on every $115 of toll. Now what ?

When a lawyer claims that out of a hundred euros of toll paid on a French highway, thirty-three end up going to shareholders in the form of dividends (once the debt is repaid, the operating charges paid and the taxes levied), you can shrug and see it as a militant approximation. Except that the figure comes from an op-ed by attorney Christophe Lèguevaques published in Les Échos in late February 2026, and that it lines up with the conclusions of three public institutions that are anything but polemical: the Senate, the General Inspectorate of Finance and the Court of Auditors.

The anomaly is documented. The mechanism is known. And yet, for twenty years, nothing moves, neither toward a renegotiation of the contracts, nor toward a renationalization. It’s precisely this inertia that a lawsuit is now trying to crack, by attacking not the ideology of privatization but its technical legality.

A rent built on €14.8 billion

The most widespread political narrative attributes the privatization of the highways to Dominique de Villepin and his Finance Minister Thierry Breton, who did indeed sign the transfer decrees in the summer of 2005. The reality is more diluted. The opening of the highway companies’ capital begins as early as 2002 with the partial IPO of ASF by Laurent Fabius, continues in 2004 with APRR under Gilles de Robien, and ends in 2005 with SANEF under Nicolas Sarkozy. The total transfer decided by the Villepin-Breton duo is merely the epilogue of a movement begun under three successive governments, of the left as much as the right.

The State then sells its holdings for about €14.8 billion, transferring to the new shareholders a debt of €16.8 billion. The buyers are three groups every French driver knows: Vinci (which grabs ASF, Escota and Cofiroute), Eiffage (which takes APRR and AREA via the Eiffarie consortium) and Abertis (which seizes SANEF and SAPN). Together, these three groups today operate more than 90% of the 9,100 kilometers of concession highways in France.

Let’s be honest about what the operation made possible: the State did pocket nearly fifteen billion euros that lightened its balance sheet in a tight budget period, the concession-holders have massively invested in the infrastructure since 2006, and the French highway network today ranks among the densest, best-maintained and safest in Europe. No one is proposing a return to the semi-public companies of the 1990s, whose management was far from exemplary. The problem isn’t that privatization was an execution failure: it’s that it was structurally miscalibrated from the start, and that the successive public reports documented this anomaly without anything being corrected.

The Delahaye Senate report of September 2020 already quantifies the shortfall of this operation: €7.8 billion in potential revenue lost by the State in 2020 value, including €1.2 billion attributable to the undervaluation of ASF: Vinci already held 23% of the capital, which discouraged competing bids and allowed the group to acquire the rest at a knockdown price.

The proof in the figures

If the 2005 operation had been defensible financially (a State parting with an asset at fair value, freeing up capital to invest elsewhere), the argument might collapse. Except that the successive public reports document the opposite, with a rare convergence between institutions.

The 2020 Senate report concludes that the profitability expected by shareholders will be reached as early as 2022, that is, ten to sixteen years before the end of the concessions, which run until 2031-2036 depending on the company. Everything the concession-holders collect between 2022 and the end of the contract therefore constitutes excess profit relative to the economic model presented to the State at the time of privatization. The independent analysis commissioned by Vincent Delahaye estimates this excess profit at about €40 billion in cumulative dividends over the 2020-2036 period, including €32 billion for Vinci and Eiffage alone.

A year later, in February 2021, the joint report of the General Inspectorate of Finance and the General Council for the Environment and Sustainable Development confirms and aggravates the diagnosis: the internal rates of return projected for shareholders at the end of the concession would reach 11.77% for Vinci and 12.49% for Eiffage, that is, far more than what had been presented as reasonable profitability at the time of the transfer. By way of comparison, the average annualized return of a diversified equity portfolio over the long term hovers around 7-8%. French highways therefore offer their shareholders a return 50% higher than that of a classic stock portfolio, for an activity that the Competition Authority was already noting in 2014 carries very low risk.

It’s precisely this last sentence that sums up the nature of the operation. The Competition Authority’s 2014 opinion put it unambiguously: “a high rate of return for a low risk is characteristic of a rent”. Not of an industrial activity, not of entrepreneurial risk-taking. A rent.

The two defense arguments, and their limit

The concession-holders counter these figures with two legitimate arguments that have to be taken seriously rather than waved away.

First argument: “We took on €16.8 billion of debt in 2005, and the dividend is only a residual after repayment and investment.” That’s true in gross accounting. On €100 of toll, the €33 paid out in dividends by Vinci are indeed collected only after payment of operating charges, debt service, taxes (the State collects about 41% of the toll price in VAT, the territorial development tax and the state-domain fee) and depreciation. The dividend is mathematically the balance, not the gross.

The argument would collapse if taking on the debt had constituted a real risk. Except that the 2021 IGF/CGEDD report documents precisely the opposite, and this is the central element the concession-holders prefer to keep quiet: the excess profitability of ASF-Escota and APRR-Area is explained essentially by the refinancing gains on their debt, a consequence of the historic fall in interest rates between 2006 and 2022. In other words, the debt transferred to the concession-holders in 2005 cost them far less to carry than expected, because monetary conditions turned in their favor: a financial windfall, not a management feat. The Treasury’s Directorate General confirms this analysis via its MARIA tool. So it’s indeed the contractual framework that allowed shareholders to capture this refinancing gain in full, with no sharing clause with the State or with users, even though the 2020 Senate report and the IGF’s own recommend the systematic introduction of such clauses in future concessions.

Second argument: “The French highway network is among the best-maintained and safest in the world.” That’s also true, and there’s no question of contesting it. The mesh is dense, the surfacing high-quality, the signage excellent, the mortality statistics among the best in Europe. The concession-holders genuinely invest in the infrastructure and deliver a service superior to the European average.

But that’s precisely the argument most easily turned around. No one is asking for less safe highways in order to pay less: the question is whether the quality of the service justifies the level of the rent extracted. The IGF/CGEDD report answers clearly: no. The maintenance and investment costs are already built into the profitability calculation, and it’s beyond these costs that the documented excess profitability sits. Better still: the mission estimated that the shareholder IRR of ASF-Escota and APRR-Area would have to be realigned with the 7.67% level targeted at privatization, which implies a substantial reduction of the rent without touching a single euro devoted to maintenance. The quality of the network isn’t the subject. It’s the argument the concession-holders use to make people believe one must choose between rent and safety, when the two are perfectly separable in their own accounting.

The monopoly that should never have been privatized

Transport Minister Gilles de Robien, in office at the time of the transfer, had opposed it internally. His phrase, uttered at the time and then repeated before the Senate inquiry commission in 2020, deserves to be pondered: “You never privatize a monopoly.”

The concession highway isn’t a market. The user who needs to link two points of the territory has no alternative highway; they have the choice between paying the rate set by the concession-holder or taking the national road, a choice that multiplies travel time by two or three. The price elasticity of highway demand is therefore very low: in other words, motorists keep paying even when the rate rises, because they concretely have no other option. This captivity of demand is a technical feature of the natural monopoly, and it’s what automatically turns any private concession into a value-extraction machine.

The traffic risk, supposed to justify the shareholders’ compensation, is also near zero. The network was already mature in 2006, traffic was growing steadily (3.7% a year on average between 1990 and 2013), and all the projections indicated the continuation of this trend. Selling a mature, captive infrastructure with inelastic demand, in a geographic monopoly, to private operators, was to guarantee their rent, not to weight it by a real risk.

This error isn’t specifically French and it isn’t specific to highways. It stems from a broader methodological question I’ll treat separately: can you cleanly separate the owner of an infrastructure from its operator, in sectors where the former cost decades of public investment and where the latter inherits a risk-free rent? The SNCF, Enedis, the highways all pose the same question. And on an even vaster scale, it’s part of a France of orders where organized capture has replaced republican redistribution. But the question goes beyond this article.

Why renationalization is a trap

The citizen’s natural reflex when faced with these figures is simple: let’s take them back. Let’s renationalize. The logic holds, in principle. It runs into two obstacles that the defenders of this solution underestimate.

The first is financial. The cost of an early termination of the concession contracts, in other words buying back the operating rights from the current shareholders, was estimated at between €40 and €50 billion in 2014, and described as “prohibitive” by the Senate report itself. Since then, the valuation of the concessions has risen with the explosion of profits, and the updated estimate now sits in a comparable, even higher range. It’s the equivalent of several major budget lines of the State, to be disbursed to recover an asset that will return to the public fold anyway in 2031-2036.

The second is operational. Over twenty years, the State has lost part of the technical competence needed to operate the network directly. You’d have to rebuild a highway administration, recruit, train, structure, all that to manage an asset that will soon be handed back. The argument is unpleasant because it implicitly validates the degradation of state capacity, but it’s factually accurate.

Renationalization is therefore not the short-term answer. It can be discussed for the post-2036 period, when the concessions reach their term and the State can choose between public management, a concession put back out to competition, or a hybrid model. For the decade now opening, the question is different: how to limit, starting today, the extraction of the rent?

The legal argument that can change everything

This is where the angle developed by the firm of attorney Christophe Lèguevaques comes in, and it’s this angle that changes the nature of the debate. His arguments don’t bear on the political advisability of privatization, but on the technical legality of the rate-setting mechanisms.

First argument: the automatic indexing of prices to inflation is, in French law, prohibited as a matter of principle. The Monetary and Financial Code proscribes clauses that mechanically tie a price to a general index. An exception was indeed introduced for highway tolls, but with no precise framing, which, according to the lawyer, legally weakens the annual rate-increase orders issued for years. The economic effect of this indexing is, moreover, deeply asymmetrical: the concession-holders’ revenue is mechanically inflated by inflation year after year, while the majority of their fixed costs absolutely don’t follow the same curve. The infrastructure was built twenty to fifty years ago and no longer has to be amortized; the debt was refinanced at historically low rates during the 2010-2020 decade; the payroll has been compressed by the automation of tolls. In other words, inflation enriches the concession-holder without impoverishing its operating balance sheet. It’s the very essence of an iniquitous rent: a variable supposed to represent the erosion of the user’s purchasing power becomes, through the contractual mechanism, the engine of the shareholder’s enrichment.

Second argument: the disproportion between the amount of the tolls and the real cost of the service rendered. The concession-holders are theoretically required to produce detailed investment inventories justifying the increases. Lèguevaques considers that these obligations are “insufficiently respected”, and that the gap between what the user pays and what the service actually costs to produce constitutes a contestable form of overbilling.

Third argument: the successive amendments to the concession contracts (notably the 2015 protocol signed under Ségolène Royal and Emmanuel Macron) extended the operating period and authorized rate increases in exchange for investments that the 2020 Senate report established were “favorable to the highway companies” and negotiated under “opaque” conditions.

The announced procedure will unfold in two stages. A formal notice to the Ministry of Transport in the summer of 2026, demanding the withdrawal of the 1995 decree on automatic increases and the setting of fair rates. In the event of refusal or no response within two months (the only credible scenarios), referral to the Conseil d’État in the fall of 2026 (France’s highest administrative court) to have the decree and the annual orders annulled. If the court recognizes the illegality, users will be able to claim compensation for their loss, which could reach up to 58% of the sums paid over the last five years according to the lawyer’s calculations based on the IGF reports.

The concrete lever, its conditions and its limits

The class action named Péage Autoroute is brought by the platform MyLeo, which specializes in mass litigation (the names of other proceedings it runs, such as the Takata airbags, Stellantis’s PureTech engines or the chlordecone scandal, give an idea of the seriousness of the setup). Registrations have been open since 2 April 2026 and close on 30 June 2026.

Three concrete conditions deserve to be stated bluntly.

First condition: being an electronic-toll subscriber. The action is strictly reserved for users holding an active subscription (Fulli, Bip&Go, Ulys, etc.). The reason is practical: only the monthly statements make it possible to constitute reliable documentary proof usable at scale. Paper toll tickets, heterogeneous and fragile, can’t be aggregated into a file that will potentially gather several hundred thousand claimants.

Second condition: paying the procedure fees. Registration costs €36 for an individual and €720 for a professional. These sums fund the fees of the lawyers bringing the action. It’s up to each person to assess whether the stake is consistent with what they spend annually on tolls: for a regular user, the registration pays for itself quickly if the procedure succeeds; for an occasional user, the investment is more symbolic.

Third condition: accepting uncertainty. No procedure before the Conseil d’État guarantees its outcome. Lèguevaques estimates the total duration at between six months and two years, and the possible compensation will come afterward, after a second litigation phase against the concession-holders if they refuse an amicable settlement. It’s a long procedure, whose final result remains suspended on the decision of an administrative court that has, in the past, already validated contested highway arrangements.

But the determining factor isn’t the guarantee of victory: it’s the fact that no other lever exists. The parliamentary reports have been piling up for fifteen years with no contractual effect. The successive ministers have given up renegotiating on substance. The concessions run until 2036. The judicial route is, to date, the only real pressure being exerted on the concession-holders and on the administration that covers them.

The comic twist that sums it all up

A small personal admission, and it speaks volumes. My electronic-toll subscription is taken out with Fulli, an operator I chose because it was the cheapest in its category. Fulli is, according to its own legal notice, “a brand of the companies APRR and AREA”, that is, of the two concession-holders owned by the Eiffage group, one of the two main players targeted by the MyLeo action.

In other words, every month I pay a subscription to the subsidiary of the group I’m about to take to court to obtain the reimbursement of its overbilling. And I can’t do otherwise, because all French electronic-toll operators are more or less tied to one of the three historic concession-holders. The user has no neutral provider. It’s exactly the situation Gilles de Robien described in 2005 under the name “monopoly”.

This absurdity isn’t a detail. It’s the most concrete possible demonstration of the fact that the competition promised by the defenders of privatization exists nowhere in the value chain: neither at the highway level (a single concession-holder per section), nor at the badge level (the subsidiaries belong to the concession-holders), nor at the fuel level (the service areas are also operated by the concession-holders, as the episode of the paid toilets at Berchem incidentally recalled). The whole ecosystem is locked down by the same three players. It’s the same mechanism that turns urban parking into an institutionalized rip-off: a captive user, a single operator over a given perimeter, and a pricing structure that no longer has anything to do with the cost of the service rendered.

What’s at stake until 30 June

The class action won’t solve the underlying problem. It won’t nationalize the highways, won’t rewrite the contracts, won’t replace the necessary political debate on what should become of the concessions at their term. But it can, for the first time in twenty years, inflict a real legal cost on a system that has never suffered one.

If the Conseil d’État recognizes the illegality of the indexing mechanism, it’s the rate foundations themselves that collapse. If compensation is granted to users, it’s the concession-holders’ balance sheets that are dented over the last five years, for amounts that will run into the billions. And if the State is forced to revise the 1995 decree, it’s the entire legal architecture of the rent that has to be rebuilt.

The chances of success are neither nil nor secured. But the information asymmetry that held for two decades (concession-holders armed with lawyers, an under-equipped State, atomized users) reverses for the first time when hundreds of thousands of subscribers band together behind a specialized firm, armed with the three public reports that prove them right.

To register, it’s here: myleo.legal/fr/products/peage-autoroute. Deadline: 30 June 2026. Cost: €36 for an individual. Required documents: ID and access to your electronic-toll subscriber account. The rest (the opportunity to have it recognized that for twenty years, we paid for a rent more than for a service) is up to each person.

I’m going for it. With my Fulli badge. Issued by Eiffage.