France at a standstill : neither fast nor at a reasonable cost

On 29 April 2026, La Dépêche announced, in the calm tone of an administrative news-in-brief, that no bank transfer would be processed in France for three days. Three days, in 2026, in the country that invented the carte bleue. Three days of shutdown on the silent backbone of the real economy, and no one is really moved by it. We file it under inconveniences. We shrug. We wait for Monday.

Then you re-read the dispatch, and the question rises. Is this normal? Twenty years ago, would anyone have dared announce this to the French as a mere maintenance operation? I have no memory of it. I have no memory, in the nineties or the two-thousands, of a French banking system shutting down over a long weekend for an update. I have no memory of a regional bridge closed for two years for renovation. I have no memory of metro escalators out of service for eleven months.

And yet, in 2026, these situations are no longer accidents. They have become the cruising speed.

Five examples that, side by side, stop being coincidences

Let’s take the most recent outages, picked almost at random from the last month’s news.

Bank transfers suspended for three days. Announced in late April by La Dépêche. The real cause is more prosaic than a “technical migration”: a bank-holiday weekend coupled with the scheduled closure of TARGET2, the European clearing house. But the technical reason hardly matters; what’s striking is the tone in which it’s announced. The implicit message to the French: make your arrangements, do your shopping, pay your rent in advance. Money won’t move over a long weekend. An outage of the national payment system has become, in 2026, an acceptable administrative news-in-brief.

The Eymeux suspension bridge, in the Drôme. Closed for renovation. Announced duration: two years. Cost: five million euros. I’m no bridge engineer, but we’re entitled to ask the naive question: how long, and how many euros, would it take today to build a new bridge? Eymeux isn’t the Golden Gate. It’s a rural structure over a modest river. And yet, two years and five million.

The Châtelet moving walkways. France 3 Île-de-France reports that three moving walkways at Europe’s largest interchange hub stayed out of service for nearly a year. On a node that hundreds of thousands of people cross every day. Eleven months to repair escalators. In Paris. In 2026.

The water and sewerage networks in the Isère. Le Dauphiné announces a large-scale municipal loan for the works to come. Once again: no technological novelty, no visionary project, just the upkeep of the pipes. And a loan is needed for it.

And a fifth, which I didn’t have to look for in the press, because I lived it. In early April 2026, Saint-Marcellin and its surroundings were cut off from the internet. Not just fibre subscribers: ADSL subscribers too. Thousands of households and businesses in the area cut off from the world. The outage lasted ten days in my case, more than two weeks for some. The structural fragility I had documented a few weeks ago regarding the local fibre, seven subcontractors for one connection, no final accountable party, ended up producing, in April, its logical consequences. The reactions on social media weren’t long in coming, on the scale of the sense of abandonment produced by two weeks without a connection in 2026, when school, remote work, banking and administrative procedures have become structurally online.

The episode brought two illuminating mechanisms to the surface. First, some operators didn’t hesitate to tell my tenants that the copper ADSL network was permanently out of use and that they absolutely had to subscribe to fibre to get the internet back: an opportunistic sales injunction, disguised as technical fate, taking advantage of an outage to sell a migration. Second, the local 5G antenna (reference 667849) having been hit by the same cascade, I found myself for 24 to 48 hours with no possible connection at all, neither fibre, nor ADSL, nor mobile. No plan B, no useful redundancy. Three technologies, a single point of failure.

I ended up taking out a Starlink subscription. When the national infrastructure fails at regular intervals, the citizen doesn’t revolt: they exit. And they exit by whatever route they can, in this case, through an American SpaceX satellite. This is the silent but massive trajectory of a growing part of the country: paying twice to obtain a service that the State, the operators and the local authorities are collectively incapable of guaranteeing.

The million euros, the new municipal unit of measurement

What’s striking in this accumulation isn’t only the duration, problematic enough in itself, but the cost.

A million euros has become the minimum unit at municipal scale. Renovating a roundabout, a new playground, bringing a gymnasium up to standard: we are now structurally above the million mark. A billion, for its part, has become the national unit: no major project escapes it. The smallest sovereign programme, the smallest large-scale public contract, the smallest sector crisis is now counted in billions.

This cost inflation follows no objective measure. The cumulative inflation of the past twenty years doesn’t explain why a rural bridge costs five million to renovate, nor why escalators take eleven months to be replaced. And to be convinced of it, you need only look at what France was capable of in eras when its technical tools were nonetheless less sophisticated.

The Millau viaduct, the tallest structure of its kind in the world, was built between 2001 and 2004, three and a half years, in a valley of sharp relief, under severe climatic constraints. In 2026, we renovate a country suspension bridge in two years for five million. The productivity ratio has become absurd. As I wrote about the providential imposture, the welfare state built nothing: it merely plundered, and when you spend forty years managing an annuity rather than maintaining a capacity, you logically arrive at this stage.

Anatomy of a paralysis

Why are we here? Before listing the pathologies, let’s concede an honest point: part of the slowness stems from a real, and largely legitimate, change in our relationship to risk. In 1980, when an escalator was put back into service, we weighed usefulness against danger by accepting that, at the margin, an accident could happen. Today, we operate on zero risk: we have to prove, document, certify, requalify the structure, train the staff, validate compliance with a framework that didn’t exist forty years ago. This raising of the safety standard is, in itself, a civilisational victory. No one wishes to return to the road tunnels of 1985 or the rough-and-ready electrical installations of the social housing of the time.

But legitimate zero risk isn’t enough to explain what we observe. We can admit that an escalator today requires more procedures than in 1990, without accepting that it should take eleven months to get it running again. Alongside this justified inflation of standards, four properly pathological causes have settled in, which reinforce one another.

1. Cascading subcontracting dilutes responsibility and multiplies the margin

When the State or the municipality places an order, the holding company no longer does the work itself. It subcontracts. And the subcontractor subcontracts. And so on. I had documented the concrete case of fibre in Saint-Marcellin: to connect one household, seven companies follow one another, each taking its margin, none bearing final responsibility. The result is arithmetical: at each link, the cost rises and the quality falls, because the principal is further and further from the executor.

What’s striking in this case, and what recurs everywhere, is the central paradox of our administrative age: the more we multiply the controls, the less anyone is responsible. In Saint-Marcellin, when a connection is botched, no one is at fault. Everyone respected their specifications, everyone ticked their boxes, everyone produces their certificate. The procedure was followed. But the result is faulty. This dissociation between procedural compliance and effective responsibility is one of the great perverse inventions of the system: it lets everyone exonerate themselves by taking refuge behind the norm that was followed. Control has replaced commitment, and commitment alone produces quality.

This model, generalised across public works, rail maintenance, the renovation of engineering structures, mechanically explains why a simple technical operation becomes a multi-year, multi-million project.

2. The administrative layering drowns the decision

We vote for the mayor, but it’s the inter-municipal authority that decides. The department arbitrates, the region co-finances, the State validates, the relevant agency issues the environmental authorisation, the architecture council gives an opinion, and the final decision falls to a steering committee where no one really has the authority to settle anything.

To renovate the Eymeux bridge, how many bodies had their say? How many preliminary studies, how many opinions, how many cross-validations? The diffraction of responsibilities has a direct cost: it multiplies the delays, because at each step a new layer has to be convinced, and it multiplies the costs, because each layer produces procedure, consultants, design offices.

And each layer doesn’t merely slow things down: it levies its tithe of time. Cumulated across an entire project, this tithe becomes prohibitive. In a world economy that’s accelerating, French administrative time has become an invisible but massive tax on reality, a tax no one pays in cash, but that everyone pays in obsolescence, in lost competitiveness, in infrastructure we won’t have had time to build before it’s already outdated.

3. The privatisation of profitable assets, the socialisation of the burdens

Let’s first clarify what isn’t the debate: privatisation, in itself, isn’t an evil. It has sometimes brought prices down and improved service, French mobile telephony in the two-thousands being its best-known illustration. The real problem isn’t privatisation. It’s the systematic asymmetry with which it was applied: we handed over the profitable flows, we kept the burdens.

The motorways are the archetype of it: 33 euros of dividends out of every 100 euros of toll, a resource the State once reinvested in the secondary network and that today funds shareholders. Meanwhile, the upkeep of departmental roads and engineering structures remains the taxpayer’s burden. The annuity leaves, the bill stays.

Direct consequence: a loan is needed to maintain the water networks, because the annuity that could have co-financed those works was privatised twenty years ago. The investment capacity of local authorities has contracted as the structuring resources slipped away from them. That’s why a municipal loan becomes the only option to renovate pipes that should have been done as a matter of course.

4. Bureaucracy devours the project before it begins

The environmental assessment, the impact study, the public procurement procedure, the consultation of local residents, the public inquiry, the water-law declaration, the preventive archaeological assessment, the opinion of the Architect of the Buildings of France if the structure is within a protected perimeter: before the first digger breaks ground, two years can have gone by. And each procedure produces its own dossier, its own consultants, its own appeals.

This bureaucracy isn’t neutral. It has a direct economic cost, and a cost in competence: the State’s best engineers no longer design, they draft specifications. The case of lumpy skin disease in the Ariège, where the administrative procedure killed an entire herd, is an extreme case, but it reveals a general mechanism: procedure has replaced decision.

The multipliers we didn’t list

We could, without changing the diagnosis, lengthen the list of factors: the rigidity of labour law that turns the slightest hire into a legal risk, the cost of dismissal that pushes towards outsourcing rather than internalising skills, the chronic shortage of qualified labour in construction, a shortage no migration or training policy has seriously tackled in twenty years. These factors aren’t the heart of the problem. But they are its multipliers.

The Eymeux bridge is a symptom, not an exception

These four causes, taken in isolation, might pass for one-off malfunctions that a good reform would correct. But together they sketch a coherent landscape: that of a country that has lost its capacity to do.

This loss isn’t accidental. It’s the cumulative product of political choices spread over three or four decades: systematic outsourcing, deindustrialisation, financialisation, regulatory stacking. And it shows up everywhere, at every scale.

At the sovereign scale, it’s the inability to see through a sovereign defence programme: SCAF, IRIS², AI, France would rather regulate its decline than build its future. At the political scale, it’s chronic instability: eight prime ministers in eight years, a turnover that would have been branded republican instability under the Fourth Republic and that has become the normal regime under the Fifth. At the budgetary scale, it’s the wall we’ve been pushing back for ten years: is mortgaging the future still acceptable? And at the monetary scale, it’s the still more worrying horizon of a collapse of the euro programmed for 2025-2030.

The Eymeux bridge isn’t an exception. It’s a fragment of the map. The same paralysis that prevents renovating an engineering structure in under two years is the one that prevents producing a European combat aircraft, stabilising a government, holding a budget, defending a currency.

A stationary state, and a way out

Nine years ago, I wrote here a cry of rage for a France running out of breath. I realise today, re-reading that 2017 text, that nothing has been corrected. The breath has just grown shorter.

France is no longer a country that does. It has become a country that manages, that manages the waiting, that manages the shortage, that manages the delay, that manages the degradation. This management has a respectable administrative name: we call it modernisation, consultation, contemporary complexity. But let’s call things by their name: it’s a slide. Not a crisis, not a cycle, not a bad patch to get through. A slow, methodical slide, of which every banking outage, every closed bridge, every out-of-service escalator is the miniature trace.

Then the question remains: what is to be done?

The answer probably isn’t in a thirty-first national reform, added to the previous thirty. It’s in the opposite movement, that of local doing, of the short circuit of decision, of taking technical matters back in hand. Shorten the chains: fewer layers between the one who decides and the one who digs. Re-internalise skills: we spent thirty years outsourcing what should have been kept in-house, we have to resolve to reverse the movement, even at the price of a rebuilt technical civil service. Rehabilitate imperfect doing against perfect procedure: a bridge renovated in eight months with a reasonable guarantee is worth more than a multi-year, zero-risk project that paralyses an entire valley.

The municipalities that still manage to carry out projects in a few months and at controlled cost do exist. They aren’t richer than the others. They’ve simply kept a mayor who decides, competent technical services, and a reasonable number of interlocutors between the deliberation and the execution. This path holds nothing seductive for the media. It doesn’t make the front page. It isn’t counted in billions. But it’s probably the one that will decide, in ten or fifteen years, whether France remains a country that does, or whether it ends up becoming a museum of its own infrastructure.

The day it’s announced, in the tone of a news-in-brief, that the water will stop flowing for three days in some department for maintenance, no one will be surprised any more. That, precisely, is the problem, and that, precisely, is what we must refuse to get used to.